






































OCR Document


 

RIDDLE USE AND COMPREHENSION 

IN IRISH SCHOOL-AGED CHILDREN: 

A DEVELOPMENTAL STUDY 

SARA STAUNTON 

I would like to gratefully acknowledge the help I received from the 
following people in completing this project: My supervisor, Dr. Jeffrey 
Kallen for his constant patience, advice and encouragement. The 
teachers, parents and children who participated in this study. My 
friends and family who assisted in the completion of this project. 

INTRODUCTION 

1.1 Introduction 

Whilst language development can be described as a lifelong process 
(Obler, 1985, cited in Nippold, 1988a), the greatest part of this 
development occurs during the childhood years. The task of acquiring his 
or her native language, with which the child is faced from birth, would 
appear to be a formidable one. However, the evidence that children rise 
to this challenge with fascination and delight is be found in the numerous 
kinds of lore which have been collected from children over the past few 
centuries (Crossley-Holland, 1982; Opie and Opie, 1959). Included 
amongst these are the rhymes, riddles, jokes and songs which children 
relish in sharing amongst themselves, and which are reported to remain 
almost unchanged when passed from generation to generation (Opie and 
Opie, 1959). 

The forms of folklore which are characteristic of the early childhood 
years, such as the nursery rhyme and fairy tale, are readily shared 
between adult and child (McDowell, 1979). However, schoolchildren's 
lore (including riddles, jokes and songs) differs from early lore in that its 
content typically remains 'partly inaccessible to adults' (McDowell, 1979: 
2). It therefore serves an important social function for children, helping to 
mark them apart from adults as a social group in their own right. Forms 
of lore such as riddles and jokes are of additional interest to linguists as 
they frequently involve the use of what is termed 'figurative language' 
(Rice and Kemper, 1984). The ability to fully comprehend and make use 
of figurative language is reported to emerge only in the later years of 
childhood (Nippold, 1988a; Rice and Kemper, 1984; Winner, 1988). 



 

STAUNTON 

1.2 Riddles 

The 'true' riddle can be defined as follows (Opie and Opie, 1959: 74), 

'a composition in which some creature or object is described in an 
intentionally obscure manner; the solution fitting all the 
characteristics of the description in the question, and usually resolving 
a paradox '. 

Opie and Opie (1959), who carried out an extensive study of the lore of 
schoolchildren during the 1950s, state that many of the riddles used by 
children in this century are actually 300 years old. The observation that 
the form of these riddles changes little when passed from generation to 
generation, is illustrated in the following 'cryptic description' of a maid 
milking a cow which was posed by a 9-year-old child in Dublin in the 
1950s (Opie and Opie, 1959: 76): 

Ink ank under a bank 
Ten drawing four, 

The same riddle was recorded in a collection in 1631 in the 
following form (In A Book of Merrie Riddles, 1631, no. 59, cited in Opie 
and Opie, 1959, 76): 

Clinke, clanke under a banke 
Ten above foure and neere the stanke. 

Although little is known about the exact cultural origins of riddles 
(Green and Pepicello, 1984) it is known that their use is by no means 
restricted to the younger members of society. There is evidence that 
people of diverse cultures have engaged in riddling for many hundreds 
of years (Crossley-Holland, 1982). Indeed, riddles are a feature of the 
rituals of a number of cultures, marking periods of transition in life such 
as baptism, marriage and death (Handleman, 1996). 

The riddle can be described as consisting of two connected parts, 
the riddle image and its answer (Kangas Maranda, 1971, cited in Lieber, 
1976). The riddle image can take a number of forms, although in English 
it usually takes the question form of the language (Kallen, 1981). Cohen 
(1996) points out that a riddle im- 

8 



 

RIDDLE USE AND COMPREHENSION 

age in the form of a question has a very different function to that of a 
genuine question. A question form is ordinarily used to obtain 
information which the questionner is seeking. A riddler, however, is 
already 'in possession of some information which he manipulates the 
addressee into seeking' through the use of the riddle image (Cohen, 
1996: 295). Although in the past the emphasis amongst those studying 
riddles was on the analysis of the riddle image, Kallen and Eastman 
(1979: 426) state that 'researchers have now begun to consider the 
answer to a riddle to be of equal importance to its question'. They refer 
to the observation by Maranda (1971) that the influence which the riddle 
answer has on the riddle image can be seen in the features of style used 
in the image, such as the use of rhyme or alliteration. 

Sources of Ambiguity 

The special relationship which exists between the riddle image and its 
answer has been described by Scott (1965, cited in Kallen, 1981:3) as a 
'partially obscured semantic fit' between the two parts of the riddle. 
Whilst one answer may appear obvious on hearing the riddle image, the 
true answer is obscured in the riddle image through the use of a 'block 
element'. The definition of a block element given by McDowell (1979: 
18) is 'that deliberate grain of confusion or ambiguity' which is contained 
within the riddle image. Green and Pepicello (1984) give a detailed 
account of the types of strategy which are used to provide a 'block 
element' in riddles. They state that the ambiguity in the riddle image may 
be either grammatically- or metaphorically-based. The latter riddling 
strategy is more typical of 'true' riddles and will be described in more 
detail in section 1.2.2. 

Block elements which are of the grammatical variety consist of 
words or phrases which possess more than one underlying form but have 
the same surface form. Grammatical ambiguity may occur at the 
phonological, morphological or syntactic levels of the language (Green 
and Pepicello, 1984). Green and Pepicello (1984: 198) conclude from an 
examination of the work of Hockett (1970) that the grammatical 
strategies which are used in many riddles originate as 'accidents of 
speech', which are found to be amusing and are later utilised in a riddle 
form They cite the riddle (193), 

9 



 

STAUNTON 

What’s  the difference between a stag fleeing from hunters and a 
midget witch? 
One is a hunted stag, the other is a stunted hag, 

as an example of the speech error of metathesis used as a riddle strategy. 
Green and Pepicello (1984: 194-97) also give examples of 

riddles which make use of the three different types of grammatical 
ambiguity. At the phonological level of the language, the most 
commonly used riddle strategy is that of simple lexical ambiguity, 
whereby a single phonetic string can be said to have two unrelated 
meanings. An example of such a riddle is, 'What turns but never moves? 
Milk', Other examples of riddle strategies at the phonological level 
include the use of metathesis (referred to above) and the use of minimal 
pairs, as in the riddle, 

What’s the difference between a baby and a coat?  
One you were, the other you wear. 

At the morphological level, the riddle strategies include the use of 
homophony between the past participle morpheme of a verb and another 
word, such as in the well-known riddle, 'What’s black and white and 
read (red) all over? A newspaper', Other strategies at this level include 
the use of ambiguity between a morpheme and a whole word (e.g. 'What 
bow can you never tie? A rainbow'), the deliberate confusion of an 
independent word with a bound morpheme, or a 'pseudomorphological' 
type of strategy, whereby a whole word (e.g. 'key') is deliberately 
confused with a phonological string (' -key') which is not actually a 
morpheme, as in the riddle, 'What's the key to a good dinner? A turkey', 

Riddle strategies employed at the syntactic level of the grammar 
include an ambiguity of phrase structure, where a surface sequence of 
words can be syntactically analysed in more than one way, such as in the 
riddle, 'How is a duck like and icicle? Both grow down '. The riddle, 
'What do you call a man who marries another man? A minister', 
involves a 'transformational' riddle strategy, whereby 'two different 
underlying structures have an identical surface form as a result of their 
syntactic derivations' (Green and Pepeicello, 1984: 195). Another 
syntactic strategy plays upon the literal and figurative interpretations of 
an idiom, such as, 'What goes most against a farmer's grain? His 
reaper'. A 

10 



 

RIDDLE USE AND COMPREHENSION 

final strategy involves a mixture of devices, such as in the riddle, 'Why 
can you not starve to death in the desert? Because of the sandwiches 
(sand which is) there, ' which involves homophony between a 
morphemic construction (the plural form of 'sandwich') and another 
syntactic interpretation ('sand which is'). 

1.2.1 Metaphor 

Metaphor is not only important as a device providing the ambiguity 
necessary to many 'true' riddles, it has also been studied in its own right 
as an important example of the' nonliteral' use of language (Rice and 
Kemper, 1984). Metaphor is defined by Franklin and Barten (1988: 299) 
as 'a use of language in which a term (or phrase) customarily applied in 
one domain is transported across conventional category boundaries and 
applied in another'. In order to comprehend nonliteral uses of language 
(such as metaphors and idioms) the listener must be able to go beyond 
the surface meaning of an utterance and determine what is referred to as 
the 'speaker meaning' (Searle, 1979). This is the meaning which the 
speaker wishes to convey through the use of the utterance. In the case of 
nonliteral uses of language, this will always be different to the surface 
form of the words uttered by the speaker. 

Lakoff and Johnson (1980) give a detailed account of the many 
ways in which metaphor is used in everyday conversation. They are of 
the view that some metaphoric expressions have become such an 
everyday part of language use that speakers of a language may not even 
be aware that they are using metaphor. For example, we talk of 
arguments as 'wars' or of time being 'in flight'. Not only is metaphor an 
important feature of language use, it therefore structures the way in 
which we think and act from day to day. The distinction can be made 
between literal and figurative metaphors (Lakoffand Johnson, 1980). 
Literal (or frozen) metaphors, such as those mentioned above, are those 
that can be seen as 'dead'. They have been assimilated into the language 
through overuse and have therefore lost their 'metaphoricity' (Winner, 
1998: 17). Figurative or novel metaphors, on the other hand, are those 
which can be seen as 'alive' as they are creative and serve to illustrate 
new ideas. 
 Richards (1936, cited in Kittay, 1987) described all metaphors 

11 



 

STAUNTON 

as having both a topic and a vehicle which are linked by a common 
ground. The topic is what the metaphor is about (i.e. the subject of the 
metaphor). The vehicle is the means through which the speaker can refer 
back to the topic. The ground is made up of the properties which link the 
topic and vehicle. An example used to illustrate these elements is 
Shakespeare's copula metaphor, 'Juliet is the sun' (Skinner, 1957, cited 
in Billow, 1977: 84). The topic of this metaphor is 'Juliet' and the vehicle 
is 'the sun'. The ground is composed of the aspects of the sun, such as its 
warmth and brilliance, which are shared by both Juliet and the sun. 

Types of Similarity 

There are a number of types of ground (similarity) on which a metaphor 
can be based (Winner, 1988). The distinction can be made between 
sensory and nonsensory kinds of similarity between the topic and the 
vehicle. Sensory metaphors are those based on similarities of 'shape, 
sound, colour, motion, texture, smell or some combination of these' 
(Winner, 1988: 64). For example, a similarity of colour is involved in the 
sensory metaphor, 'The bird was a rainbow flying in the sky '(Nippold, 
Leonard and Kail, 1984: 198). 

Nonsensory metaphors are grounded in the types of similarity which 
are not directly perceived by our senses. This similarity may be of a 
relational or a psychological-physical nature. Metaphors based on a 
relational type of similarity link the topic to its vehicle due to a similarity 
in the way in which the two items function. An example is the likening 
of the human brain to a computer. The suggestion here is that the brain 
processes information in the same way a computer does (Winner, 1988). 
Other metaphors in use in everyday language, such as the notions of time 
'flying' or ideas 'blossoming' (Lakoff and Johnson, 1980), involve similar 
types of ground. 

Psychological-physical metaphors are usually based on a similarity 
between 'sensory attributes of a physical object. . . and psychological, 
nonsensory attributes of a person' (Winner, 1988: 67). Included here are 
dual-function terms, such as 'warm', 'sweet' or 'soft' which can be used to 
describe the physical attributes of objects or the personality traits of 
people (Asch and Nerlove, 1960, cited in Billow, 1975). Asch (1958, 
cited in Billow, 1977: 85) gives 

12 



 

RIDDLE USE AND COMPREHENSION 

the example of the adjective 'hard', which could be applied in a physical 
sense to a piece of furniture or in a psychological sense to a person. 
When used to describe a table, the term 'hard' means that the table is 
resistant to change when physically pushed or pressed. To describe a 
person as 'hard' might imply that he or she is unlikely to yield to pressure 
from others. The psychological application of the term is therefore 
linked to its physical sense, through the notion of an entity resisting 
change when force is applied. Asch (1955, 1958, cited in Billow, 1977) 
also reports that historically unrelated languages have been found to 
contain the same metaphoric descriptions of personality traits, indicating 
that this is not an arbitrary phenomenon. 

1.2.2 Metaphoric Riddles 

Good riddles do, in general, provide us with satisfactory metaphor: 
for metaphors imply riddles and therefore a good riddle can furnish a 
good metaphor. (Aristotle, trans. Rhys Roberts, cited in Hackiyan, 
1966: 2) 

A 'true' riddle can be said to involve a metaphoric comparison between 
two entities which is obscured in some way so as to puzzle the riddlee. A 
theorist who has contributed greatly to the study of such riddles is Elli 
Köngäs Maranda (1971: 196) who states that the metaphors involved in 
riddles 'are conditional metaphors, and the riddle image states the 
condition under which the metaphor holds true'. One-half of the 
metaphoric comparison is therefore revealed in the riddle image (along 
with information which obscures the answer) and the task for the riddlee 
is to complete the metaphor, thereby answering the riddle (Green and 
Pepicello, 1984). 

Based on her analysis of Finnish riddles, Köngäs Maranda (1971: 
198-99) concludes that the metaphoric riddle comprises five basic 
elements. An example used to illustrate these elements is the Finnish 
riddle, 'One pig, two snouts', to which the answer is 'A plough' (referring 
to the traditional Finnish 'fork plough' which has two prongs or 'snouts'). 
The first element of the riddle is labelled the' signans' or given term of 
the metaphor (in this case, 'a pig'). Secondly, there is a constant premise 
(similar to the 'ground' 

13 



 

STAUNTON 

of the metaphor described by Richards (1936, cited in Kittay, 1987) 
which is true for both the signans, and the the 'signatum'(riddle answer). 
In this case, the constant premise is that both a pig and a plough can 
have snouts. The third element is the hidden variable which is never 
mentioned in the riddle (e.g. that a pig can only have one snout, not two) 
but which indicates to the riddlee that something which is stated in the 
riddle image needs to be resolved. The fourth element is the given 
variable of the riddle (e.g. that two snouts are involved) which is 
described as 'the condition under which the metaphor holds true' 
(Köngäs Maranda, 1971: 199). Finally, the other hidden term is the 
riddle answer or 'signatum' which the riddlee is seeking. 

Köngäs Maranda (1971) also observes from her study of Finnish 
riddles that riddles create a surprise by combining elements from 
categories or 'sets' which are normally seen as being opposites. Köngäs 
Maranda found in her analysis of seventy-two riddles that the most 
commonly occurring contrast between the categories was that of the set 
of animate entities versus the set of inanimate entities. Within these two 
sets, almost half the riddles involved a juxtaposition between human 
objects and cultural objects. About a quarter of the riddles involved the 
contrast between human beings and wild plants. Only a few riddles 
contained reference to supernatural entities such as God or to abstract 
concepts such as number and colour. Metaphoric or 'true' riddles might 
thus be said to play on listener's ability to view two entities as being 
equivalent (on a sensory level) due to their shared characteristics, despite 
the basic knowledge on an idealogical level that the two things are not 
the same and in fact belong to different classes (e.g. animate things 
versus inanimate things) (Kallen and Eastman, 1979). 

Types of Riddle Metaphor 

A distinction can be made between riddles which make use of ‘frozen’ 
metaphors for their ambiguity and those which involve the generation of 
a novel metaphorical comparison (McDowell, 1979: 95). The term 
‘conventional polysemy’ is used to describe metaphors which are 
present in everyday language use (i.e. frozen metaphors). Conventional 
polysemy occurs where the two mean- 

14 



 

RIDDLE USE AND COMPREHENSION 

ings of a particular phonetic string are thought of as being related, 
through a metaphorical comparison (e.g. the leg of a person and the 'leg' 
of a table). The speakers of the language intuitively consider the two 
uses of the word the be related, whether or not this is historically the 
case (McDowell, 1979). 

McDowell (1979: 98) cites other examples of riddles which require 
the literal interpretation of 'dead' or frozen metaphors which have been 
assimilated into the language. For example, the riddle 'Why did the boy 
throw his clock out the window?' (To see time fly), involves the literal 
interpretation of the well-known metaphor of time being in fight. A 
similar type of riddle involves what McDowell (1979: 97) describes as 
the' resurrection of automatized language' through the use of fragments 
of imaginary conversations between usually inanimate objects. These 
riddles take the form 'What did the X say to the Y?' (McDowell, 1979: 
97), such as in the riddles 'What did the rug say to the floor?' (I've got 
you covered) or 'What did the big telephone say to the little telephone? ' 
(You're too young to be engaged). The ambiguity involved in each of 
these riddles revolves around the two senses of a particular word or 
phrases, in this case the two senses of the words 'cover' and the phrase 'to 
be engaged'. 

In addition to the use of metaphors already sanctioned in the 
language, riddles can involve the strategy termed 'radical' or novel 
polysemy. Radical polysemy, as described by McDowell (1979), 
involves the use of a metaphorical comparison between two things, 
which has not already been assimilated into the language. The most 
common example is the use of metaphor to remark on the similarity 
between the sensory (physical) features of two items. The two items may 
have little in common other. than their appearance, 'but these 
incongruities are momentarily held in abeyance while the common 
features are held in focus'(McDowell, 1979: 101). 

Radical polysemy was the strategy used in many of the riddles in the 
sample analysed by Köngäs Maranda (1971). As mentioned earlier, she 
too commented on the fact that the items compared were from vastly 
different categories or sets. McDowell (1979) notes that the main types 
of comparison in use in his sample of children's riddles are those of 
culture versus nature (e.g. a cherry compared to police car lights), 
animate versus inanimate entities (e.g. an old lady compared to a 
candle), mankind versus other 

15 



 

STAUNTON 

forms of life, and finally, cultural artifacts compared with each other. 

1.2.3 Solving Riddles 

There is some debate regarding the processes involved in the solving of 
'true' riddles and in the selection of 'acceptable' answers. Kaivola-
Bregenhøj (1996:30) in a discussion based on Finnish riddles, suggests 
that riddles in which the riddle image is 'semantically vague' can 
sometimes be solved by guess-work, as there are a greater number of 
possible answers which fit. The degree of validity of suitable but 
unexpected answers seems to depend, however, on the judgement of the 
particular riddler and on the customs of the culture in which the riddling 
session is taking place. Kallen and Eastman (1979: 418), for example, 
conclude from their study of Swahili riddles that such riddles are solved 
through the use of 'symbolic activity' and interaction between the parties 
involved, rather than a process of reflection and analysis by the riddlee. 

Köngäs Maranda (1971, cited in Kaivola-Bregenhøj, 1996) remarks 
that there is a tendency amongst riddlers of accepting only their original, 
preferred answer and rejecting other suitable answers. Kaivola-Bregenhøj 
(1996:32) concludes that in a riddling session the riddler and riddlee take 
two different approaches to the solving of a riddle. The riddlee uses the 
process of 'creative guessing' in order to select a suitable answer to the 
riddle. The answer which the riddler has in mind, however, is founded on 
'a conventional interpretation of the image that does not tolerate any 
variation' . 

Whilst the validity of a particular answer may vary depending 
on the cultural context of the riddling session or the judgement of the 
riddler, it would appear that there are a number of conditions which the 
answer to a true riddle must fulfill. It has been remarked that a riddle 
answer must complete the condition described as a 'partially obscured 
semantic fit' which exists between the riddle image and its answer (Scott, 
1965, cited in Kallen, 1981: 4). In order to satisfy this condition, the 
riddle answer must share enough features with the riddle image for them 
to be seen as equivalent within the context of the riddle (through the use 
of metaphor). 

16 



 

RIDDLE USE AND COMPREHENSION 

However, the riddle answer must also fulfill the requirement of being 
significantly different in nature from that of the image, so that the 
semantic fit between the two of them is sufficiently obscured. 

The work of Elli Köngäs Maranda (1971) revealed that there are 
certain semantic groupings which are traditionally more likely than 
others to be linked in the metaphors involved in true riddles. As 
observed from her study of Finnish riddles, true riddles generally involve 
the juxtaposition of opposites (Köngäs Maranda, 1971). The item which 
appears to be described in the riddle image typically belongs to an 
opposite semantic category or 'set' to that of the riddle answer. Aitchison 
(1994: 150) also comments that in a good metaphor, 'the subject of the 
metaphor is compared to something which is quite different from itself, 
in the sense of coming from a different semantic field'. Thus, the first 
criteria which a riddle answer might be required to meet is that of 
belonging to a semantic category which is opposite to that of the riddle 
image. 

Secondly, the notion put forward by Rosch (1975), that some 
members of a category are considered more 'prototypical' examples of 
category members than others, might also be borne in mind when 
judging the degree of obscurity of fit between the riddle image and a 
possible answer. A riddle answer, one might assume, should be a 
prototypical example of a member of a category opposite to that item in 
the riddle image, if the riddlee is to be given the chance to 'solve' the 
riddle on the first hearing. 

A third and essential feature of the riddle metaphor is that the item 
named the riddle answer must have characteristics in common with the 
item mentioned in the riddle image. Aitchison (1994: 150) remarks that 
although the items compared 'must not share major characteristics, they 
must share' some'. She states that good metaphors typically involve a 
comparison based on a minor, but 'fairly obvious' characteristic. This 
third condition, along with those mentioned above, must be fulfilled 
therefore, if a particular answer is to be judged as completing the riddle. 

1.3 Children's Riddling 

The most comprehensive developmental account of children's riddling in 
the literature is that given by John Holmes McDowell (1979). 
McDowell conducted a detailed analysis of the riddles 

17 



 

STAUNTON 

collected from a sample of fifty working class, chicano children and 
twenty-five anglo, middle class children living in Austin, Texas, during 
the 1970s. The children ranged in age from 4 up to 11 years, though 
McDowell's analysis focuses mainly on the riddles used by children 
between the ages of 5 and 8 years. 

McDowell (1979: 31) uses the term 'interrogative ludic routine' to 
label all playful linguistic sequences (such as riddles) which involve an 
inversion of the everyday speech act of interrogation. He reserves the 
terms 'riddle' or 'riddle proper' to refer only to interrogative ludic 
routines which contain an identifiable block element. McDowell found 
from the analysis of his corpus of riddles that there is a distinct 
developmental pattern in the kinds of ludic routines which children use. 

The riddling of children of 4- and 5-years-of-age is characterised by 
the use of what Sutton-Smith (1976, cited in McDowell, 1979: 32-33) 
labels 'preriddles', as children frequently do not yet show awareness of 
the relationship between the question and answer of the riddle. Children 
of this age tend either to reproduce a riddle which they have heard 
elsewhere, without actually comprehending it, or they 'create puzzling 
questions and give them arbitrary answers'. Another feature which is 
characteristic of the riddling of children at this age (and which is a 
feature of riddling at all ages) is the use of interrogative ludic routines 
which contain no block element and which are 'not intended to confuse' 
(McDowell, 1979: 34). These early routines are labelled 'descriptive 
routines' as the riddle image typically involves a description of one or 
more attributes of an object. Examples of such routines used by the 
chicano children include (McDowell, 1979: 35): 

What has five sides and lives in the sea? A starfish.  
What’s square and gots a point at the top? A house. 

Whilst most descriptive routines are based on the physical char-
acteristics of objects, they can also involve the themes of instrumentality 
(e.g. 'What do you need feet for? To walk') or causation ('How come a 
rabbit goes into a hole? Cause she eats a lot of carrots '). 

By the age of six years, children have begun to use the 'riddle 

18 



 

RIDDLE USE AND COMPREHENSION 

proper', whilst still using a large number of descriptive routines. The 
riddle proper differs from the descriptive routine in that it contains a 
source of ambiguity (the block element) and its purpose is to trick or 
confuse, rather than help the riddlee, in finding the answer. The riddle 
strategies used at this stage are primarily homophony (described earlier) 
and anomaly, which is the strategy at work in a riddle such as, 'What 
goes up and never comes down? Your age '. McDowell (1979: 106) 
states that riddles involving anomaly serve to challenge everyday states 
of affairs which we have taken for granted (such as the laws of gravity, 
in the riddle above). 

By seven years of age, children have become more adept at the use 
of the riddle proper. There is a strong preference for the use of riddles 
involving all types of linguistic ambiguity amongst children of this age. 
There is also a marked increase in the use of riddles involving the 
polysemies (metaphor) at this stage. Anomaly riddles involving the 
theme of fantasy, such as those involving the generation of 'unnatural 
facts', are present at this age (e.g. 'What’s big and red and eats rocks? A 
big, red rockeater', McDowell, 1979: 253). However, riddles involving 
'real world' anomaly are equally popular with this age group. 

The repertoire of 8-year-old children consists mainly of the riddle 
proper. McDowell (1979: 208) reports that by this age 'the riddler's 
progress appears to be virtually complete', with a more sophisticated 
execution of riddling routines. The items featured in the descriptive 
routines of children of this age tend to be more exotic than before. 
Whilst children in this age group still use riddles involving linguistic 
ambiguity, there is an overall decrease in their use of homophony as a 
block element. They also favour the use of anomaly riddles with a 'real 
world' theme. McDowell remarks that children's interest in riddling 
levels off during the ninth and tenth years, and seems to be diverted to 
other performance genres at this stage. 

Types of Riddle Structure 

In addition to identifying the types of riddle strategy which were used by 
the children, McDowell found that the riddles in his sample were 
structured in five different ways. Over two-thirds of the 

19 



 

STAUNTON 

riddles (and descriptive routines) in McDowell's sample involved the 
construction devices of attribution, causality and instrumentality. Riddles 
based on attribution, involve the use of many simple descriptive units, 
which McDowell (1979: 65) describes as being 'stacked'. Examples from 
his study include the following: 

What is white on the outside and green on the inside and hops? Frog 
sandwich. 

What’s big and red and eats rocks? A big, red rockeater. 

Riddles associated with causality usually take the form of a question 
and answer sequence involving 'Why' and 'Because'. Examples given by 
McDowell (1979: 66) include: 

Why do birds fly south? Cause it's too far to walk.  
Why did the man take hay to bed? To feed the nightmares. 

A less frequent structure used in McDowell's sample is that of 
instrumentality, such as in the riddle, 'How do you keep a skunk 
from smelling? Hold his nose '. 

The two other riddle structures which were used in the study take the  
form, 'What did the X say to the Y?' and 'What kind of X is a P '. In the former 
riddle structure, inanimate objects are 'endowed with speech' (McDowell, 
1979: 67) and the two possible meanings of a well-known figure of speech 
are exploited. A well-known example is the riddle involving an imaginary 
conversation between two chimney pots, 'What did the big chimney say to the 
little chimney? You're too young to smoke '. Finally, riddles based on the 
structure, 'What kind of X is a Y? ' appear classificatory in nature. Yet they 
frequently involve the use of conventional polysemy, whereby a dead 
metaphor must again be resuscitated in order to solve the riddle, as in the 
example, 'What kind of head grows in the garden? A head of lettuce' 
(McDowell, 1979: 68). 

1.3.1 Children's Comprehension of Riddles 

The development of the ability to comprehend riddles has been 
investigated by a small number of researchers (Fowles and Glanz, 1977; 
Prentice and Fathman, 1975; Shultz, 1974). Most of the 

20 



 

RIDDLE USE AND COMPREHENSION 

available research has examined the comprehension of joking riddles 
which make use of the types of grammatical ambiguity outlined by 
Green and Pepicello (1984). The development of the comprehension of 
'true' riddles has not specifically been examined, although the study by 
Fowles and Glanz (1977) did include some riddles involving the use of 
metaphoric language as a source of ambiguity. The development of 
comprehension of metaphor outside of a riddling context has also been 
examined by researchers and the findings of these studies will be 
described in section 1.4. 

Fowles and Glanz (1977) examined the comprehension (and 
appreciation) of riddles in a group of fourteen children, aged between 6 
and 9 years (Grades 1 to 3 at school). The children were presented with 
a series of riddles and were asked to retell and explain each of them. 
These riddles mainly involved the use of grammatical ambiguity as a 
block element. 'True' riddles were not specifically included, although 
some of the riddles contained terms which had both a literal and 
metaphorical meaning. On examination of the results, Fowles and 
Glanz (1977) report that the children in the 7-8 year age group 
performed better than those in the age groups above (8-9 years) and 
below them (6-7 years). Although they found that the level of 
comprehension was not closely related to age, the researchers report 
that riddle comprehension was related to reading ability (as rated by 
teachers) in each of the age groups. 

Fowles and Glanz (1977) also remark from their results that the 
riddles which the children found most difficult to comprehend were those 
which contained more advanced vocabulary and those containing 
abstract or metaphorical language. Riddles in which the two possible 
interpretations of the ambiguous terms were both concrete were found to 
be easier to comprehend. Familiarity with the riddles themselves 
facilitated retelling of the riddles but did not necessarily aid 
comprehension of them. The researchers state that the influence of 
factors which are thought to affect ease of comprehension (such as the 
use of metaphorical meaning as opposed to more concrete meaning) is in 
need of further investigation, 'using riddles chosen specifically to 
exemplify these polarities' (Fowles and Glanz, 1977: 450). 

In a similar study, Shultz (1974) investigated the comprehen- 

21 



 

STAUNTON 

sion of riddles in a group of thirty children in the 6-, 8-, 10- and 12-year 
age groups. An explanation task was again used to examine the children's 
comprehension of grammatically-based riddles. Comprehension of the 
riddles was found to increase steadily with age. The 6-year-olds had 
particular difficulty explaining the two possible interpretations of the 
ambiguity in the riddles. They were judged to have comprehended just 
under a quarter of the riddles, with the 8-year-olds comprehending twice 
this number. The level of comprehension increased to just under two-
thirds of the riddles amongst the 10-year-olds and the 12-year-olds 
performed slightly better again. 

Prentice and Fathman (1975) also used an explanation task to 
investigate the comprehension of joking riddles in a group of forty-eight 
boys and girls aged 6, 8 and 10 years. Three types of riddles were 
included in the study, namely those with a theme of aggression, those 
with a caring or nurturing theme and those with a neutral content. The 
influence on comprehension of different types of linguistic ambiguity 
was not examined. Prentice and Fathman report that riddle 
comprehension was found to increase almost linearly with increasing 
age. No major differences in comprehension of the different riddle types 
was found between boys and girls. Overall, comprehension of the riddles 
was also found to be significantly related to intelligence. A vocabulary 
test had been used to rate intelligence, however, which in itself is a verbal 
task and does not examine other cognitive abilities. The older children in 
the study showed greater comprehension of the riddles with an 
aggressive or nurturant content than the younger group. The researchers 
suggest that this may be due either to an increased ability to cope with 
riddles with an affective content as the child matures or to a difference in 
the cognitive demands of the different riddle types. 

1.4 Development of the Use of Metaphor 

Traditionally, the study of metaphor was confined to the domains of 
philosophy and literature (Winner, 1988). Research into children's 
ability to use and understand metaphor has only come about in essence 
in the past thirty years, with increased interest in the topic among those 
working in the fields of linguistics and cog- 

22 



 

RIDDLE USE AND COMPREHENSION 

nitive psychology. There is some debate as to when children actually 
begin to use metaphor in their speech. It has been widely observed that 
as preschool children develop language, the phenomenon of 
overextension of words occurs (Bee, 1992). It is thought that children 
apply a word too broadly because they have not yet learned the correct 
name for a particular concept. Clark (1973) gives the example of the 
word 'ball' being extended to label other similarly shaped objects such 
as apples, doorknobs and balloons. Winner (1988: 90) points out that 
'overextensions, like metaphors, are based on similarity, and hence like 
metaphors, offer clues to how easily children classify the world'. It has 
been argued that some overextensions may in fact be metaphors 
(Billow, 1981). Winner (1988) is of the view that while children may 
overextend words as they lack the correct name for an object, the use of 
metaphor serves a very different function. A child who is using 
metaphor might already know the name of the object, but is using a 
metaphoric name in order to remark on a similarity which he has no-
ticed. 

Few studies have examined children's early use of metaphor, but 
those by Winner (1979) and Winner, McCarthy, Kleinman and Gardner 
(1979, cited in Winner, 1988) concluded that two kinds of metaphor are 
characteristic of the speech of young children. The earliest emerging 
metaphors appear to be those involving the renaming of an object in the 
context of symbolic play. These renamings are usually based on the 
actions which an object displays, or on the function of the object (e.g. 
the child featured in the first study put his foot inside a wastebasket and 
said 'boot').  Sensory metaphors are reported to be the second form to 
develop and occur where the child notices a physical similarity between 
two objects (usually their shape), without using them in pretend play 
(Winner, 1979). Winner (1979) also remarks that the metaphors used by 
preschool children are less sophisticated than those of adults or older 
children in that they consist solely of the renaming of physical objects 
and are uttered only in the immediate presence of those objects. The 
renamings are also based on a sensory (physical) or functional similarity 
which the child has noticed, rather than using the more abstract types of 
comparison typical of adult metaphors. 

23 



 

STAUNTON 

Development of Metaphor in School-Aged Children: 

A number of studies (Billow, 1981; Gardner, Kircher, Winner and 
Perkins, 1975; Pollio and Pollio, 1974), have suggested that the 
incidence of children's spontaneous use of metaphor decreases sharply 
during the school years. The reasons for this may include the fact that, as 
they grow older, children are less likely to engage in the kinds of 
activities (such as symbolic play) which stimulate the use of spontaneous 
metaphor (Billow, 1981). Furthermore, it has been noted that between 
the ages of eight and ten years, a literal or 'conventional' stage occurs in 
children's use of metaphor (Winner, 1988). During this stage, children 
become far less likely to use metaphor themselves, and may even be 
resistant to its use. Gardner, Kircher, Winner and Perkins (1975: 128) 
found that given a choice of endings for a sentence such as 'He looked as 
gigantic as...', children in this age group chose unimaginative, 
conventional endings above all other types of ending. They rejected 
novel endings (which were preferred by adolescents and adults in the 
study), insisting for example, that the term 'loud' could not be applied to 
a colour. Winner (1988) reports that this concern with convention is also 
reflected in the drawings of children of this age. 

Other research has suggested that the ability to use metaphor 
continues to develop during the school years, but that the type of writing 
tasks typically used in school do not encourage children to make creative 
use of metaphor. Pollio and Pollio (1974) examined the metaphoric 
productions of a group of children aged between the ages of eight and 
eleven years on a story-writing task. The children's use of both novel and 
frozen (or 'dead') metaphors decreased as their class level in school 
increased. However, the researchers later presented the same group of 
children with a structured comparison task, whereby the children were 
asked to verbalise the similarities between pairs of unrelated items (e.g. 
'can and box'). Overall, the children used a greater proportion of novel 
than frozen metaphors on this task, and their production of both types of 
metaphor increased (rather than decreased) with their class level. The 
findings of this study suggest that children in the eight to eleven age 
group continue to develop the ability to use metaphor, but that in 
completing school tasks they are increas- 

24 



 

RIDDLE USE AND COMPREHENSION 

ingly concerned about attaining good marks and so 'are less likely to 
'rock the boat' by using words in unconventional or nonliteral ways' 
(Nippold, 1988b: 189). 

1.4.1 Development of Comprehension of Metaphor 

There is controversy as to the age at which children first exhibit the 
ability to comprehend metaphor (Winner, 1988). However, most 
researchers agree that as children get older, their comprehension of 
metaphors increases. Early research into children's comprehension of 
metaphor was based on the 'cognitive prerequisite theory' of 
development (e.g. Billow, 1975; Cometa and Eson, 1978; Smith, 1976). 
According to this view, the attainment of the Piagetian stage of formal 
operational thought is necessary before comprehension of forms of 
figurative language can occur (Piaget, 1926, cited in Kamhi and 
Friemoth Lee, 1988). Thinking during the formal operational stage 
(which begins at about twelve years) is characterised by a move' away 
from the reality of the physical data of the senses towards an anticipation 
of abstract 'potential' properties of the environment' (Smith, 1976: 242) . 
It was hypothesized that full comprehension of metaphor would not be 
possible prior to the emergence of this type of abstract thinking. Meta-
phor comprehension, according to this theory, would therefore be 
expected to emerge during the late childhood and early adolescent years. 

The findings of the studies by Billow (1975), Cometa and Eson 
(1978) and Smith (1976), which examined children's comprehension of 
metaphor in relation to their Piagetian stages of development, did not 
support the cognitive prerequisite hypothesis. The children in Smith's 
(1976) study, for example, were shown to have an understanding of 
metaphor before they had reached the formal operational stage of 
thinking. In addition, a number of studies have reported the finding that 
even preschool children show a basic understanding of metaphor when 
simplified testing procedures and age-appropriate materials are used 
(Gardner, 1974; Gentner, 1977; Vosniadou, Ortony, Renolds and 
Wilson, 1984; Winner, 1979). Nippold (1988b) reports that although 
there may be some relationship between cognitive development and 
metaphor comprehension, no specific pre-requisites for metaphor 
comprehension 

25 



 

have been identified and the cognitive pre-requisite hypothesis is no 
longer generally accepted as a theory of metaphor comprehension. 

STAUNTON 

The Language Experience Hypothesis 

A theory which has been more widely adopted is the 'language 
experience hypothesis', which was proposed by Ortony, Turner and 
Larson-Shapiro (1985). According to this view, the ability to 
comprehend and use figurative language develops gradually through 
meaningful exposure to this type of language. As they grow older, 
children have more exposure to figurative language in both its spoken 
and written forms, in a variety of contexts. They also have more 
opportunity to test out their understanding through their own use of such 
language (Nippold, Taylor and Baker, 1996). The finding of a number of 
studies that the depth of comprehension of metaphor increases as 
children grow older might be explained by this theory (Asch and 
Nerlove, 1960, cited in Billow, 1977; Gentner and Stuart, 1983, cited in 
Winner, 1988; Pollio and Pickens, 1980; Pollio and Pollio, 1979; 
Winner, Rosenthiel and Gardner, 1976). 

Another implication of the language experience theory is that 
metaphors which occur more often in the language (i.e. frozen rather 
than novel metaphors) should be easier for children to comprehend than 
those which have less frequent usage. This notion was explored in 
studies by Pollio and Pollio (1979) and Pollio and Pickens (1980), 
involving children between the ages of 8 and 17 years. Both studies 
reported greater comprehension of the frozen than the novel metaphors 
at every age level, with comprehension of both types of metaphor 
increasing steadily as the children got older. 

In addition to the novelty of the metaphorical expression, re-
searchers have examined the influence on children's comprehension of 
factors such as the type of task used to measure comprehension, the type 
of metaphoric similarity involved and the complexity of the syntactic 
form of the metaphor. The findings of research into these aspects will be 
described in the next sections. 

26 



 

RIDDLE USE AND COMPREHENSION 

1.4.2 Measurement of Comprehension 

Winner (1988: 45) concludes from a review of studies into metaphor 
comprehension that the method used to measure comprehension, 
'dramatically affects the level of comprehension revealed'. The 
contradictory findings of various studies as to when metaphor 
comprehension emerges have been attributed in large part to the 
differences in the demands of the tasks used. The two tasks which are 
most frequently used to assess metaphor comprehension are multiple 
choice measures (e.g. Gardner, 1974; Nippold, Leonard and Kail, 1984) 
and explanation or paraphrasing tasks (e.g. Asch and Nerlove, 1960, 
cited in Billow, 1977; Smith, 1976; Winner, Rosenthiel and Gardner, 
1976). 

Multiple-choice tasks require children to choose from a set of 
possible meanings or pictures showing the meanings of the metaphors 
presented. However, these tasks may either under- or overestimate 
children's comprehension of the metaphors, depending on the type and 
range of options presented (Winner, 1988). Tasks in which children are 
asked to paraphrase metaphors offer insight into the degree and nature of 
their understanding (Nippold and Rudzinski, 1993). The possibility of 
over-estimating a child's level of comprehension is also greatly 
diminished. Such tasks have also been criticised, however, due to the 
metalinguistic demands involved (Vosniadou, 1987). The finding of 
some studies that metaphor comprehension only emerges in late 
childhood might therefore be attributed to the significant meta1inguistic 
demands of the tasks used (e.g. Billow, 1975; Cometa and Eson, 1978; 
Smith, 1976). 

A study by Gardner (1974) compared the performance of children 
aged between 3 and 11 years on a multiple-choice picture task and an 
explanation task. The level of accuracy using the receptive condition was 
found to be greater than with the paraphrasing task. Similarly, a study by 
Winner, Rosenthiel and Gardner (1976) involved the use of a 
paraphrasing measure and a written multiple-choice task. Children who 
had difficulty paraphrasing the metaphors were often found to have 
performed well on the multiple-choice measure. Pollio and Pickens 
(1980) reported parallel findings in their study. In evaluating and 
comparing the findings of various studies, it is therefore essential that 
the demands of the measures used be taken into account. 

27 



 

STAUNTON 

1.4.3 Types of Similarity 

A factor which is thought to have a large influence on children's 
comprehension of metaphor is the type of similarity on which the 
metaphor is based. Winner (1988) cites the findings of (unpublished) 
studies by Gentner and Stuart (1983), Mendelsohn, Gardner and Winner 
(1981) and Shantiris (1983), that nonsensory metaphors pose more 
difficulty for younger children than those grounded in sensory similarity. 

In the study by Shantiris, children of different ages were presented 
with a series of sensory metaphors, nonsensory metaphors and 
anomalies. The children were asked to decide whether statements such 
as, 'Eyes are marbles' (74) made sense, thereby indicating 
comprehension of the grounds. At all ages, children found psychological-
physical metaphors the most difficult type to comprehend. However, 
relational metaphors proved no more difficult than sensory metaphors for 
any age group. A different finding regarding relational metaphors was 
reported by Gentner and Stuart, whose study involved children aged 5-6 
years and 9-10 years. Metaphors based on a sensory similarity (e.g. 'A 
cloud is like a marshmallow') and relational similarity (e.g. 'Tree bark is 
like skin' [p. 74]) were presented to the children. When asked to explain 
the sensory metaphors, children aged 5-6 years performed as well as the 
older children. They had considerable difficulty explaining the metaphors 
grounded in a relational type of similarity. 

Mendelsohn et al. (1981, cited in Winner, 1988) used a multiple-
choice task to probe for comprehension of sensory and nonsensory 
metaphors in children aged 6 years and above. Children of all ages were 
found to have more difficulty with metaphors grounded in relational or 
psychological-physical similarity than those involving sensory grounds. 
A study by Nippold, Leonard and Kail (1984) contradicted this finding in 
relation to psychological-physical metaphors, however. Their study 
involved children aged 7 and 9 years who were again presented with a 
multiple-choice task. They reported no difference in the ease of un-
derstanding between sensory and psychological-physical metaphors. 
Kogan, Connor, Gross and Fava (1980, cited in Winner, 1988) reported 
similar findings when a basic picture matching task, with minimal 
linguistic demands was used. 

28 



 

RIDDLE USE AND COMPREHENSION 

Literal Interpretation of Nonsensory Metaphor 

Studies where explanation tasks are used offer insight into the nature of 
younger children's interpretation of nonsensory metaphor. The finding 
that younger children tend to interpret nonsensory metaphors in a literal 
(physical) sense has been commonly reported. An early study by Asch 
and Nerlove (1960, cited in Billow, 1977) investigated the 
comprehension of dual-function terms such as 'hard', 'sweet' and 'warm' 
in children aged between 3 and 12 years. Whilst the preschool children 
comprehended the dual-function adjectives in reference to physical 
objects, they showed no understanding that people could also be 
described in this way. The 5- and 6-year-olds in the study understood 
that the adjectives could be applied to people, but only in a physical 
sense (e.g. 'cold' people were understood to be people who were not 
warmly dressed). Only between the ages of 7 and 10 years did the 
children show awareness that the terms could be applied in a psy-
chological sense to people. And only the 11- and 12-year-olds (who 
understood the metaphors with a level of 96% accuracy) were 
consistently able to verbalise the metaphorical fink between the two 
senses of the terms. 

A study by Winner, Rosenthiel and Gardner (1976) also involved 
the presentation of dual-function terms, but in predicative metaphors. 
They report that 30% of 8-year-olds and 48% of 10year-olds were able to 
explain the psychological meaning of the terms. Prior to this, the children 
tended to interpret the metaphors literally (e.g. the description of a prison 
guard as a 'hard rock' was interpreted as meaning that he had hard, tough 
muscles). 

The finding that nonsensory similarity seems to pose more difficulty 
for children was explained by Keil (1979, 1985, cited in Winner, 1988) 
in terms of a lack of 'domain knowledge'. The term 'domain' is used here 
to refer to the categories by which humans beings organise their world 
(such as furniture, modes of transport or personality traits). 'Domain 
knowledge' therefore refers to the knowledge which a child has about the 
properties and functions of things within a particular domain and about 
the boundaries of that domain. Keil was of the view that in order for a 
child to comprehend a metaphor, he or she must be not only be familiar 
with the referents of the topic and vehicle, but must also have knowl- 

29 



 

STAUNTON 

edge of the two domains to which the topic and vehicle belong. Keil's 
hypothesis that knowledge of the world increases as children get older 
was supported by the findings of the study by Gentner and Stuart (1983, 
cited in Winner, 1988) mentioned earlier. Prior to the presentation of the 
metaphor task used in their study, Gentner and Stuart asked each child to 
describe the items mentioned in the metaphors (e.g. tree bark, skin etc.). 
The list of properties which the children gave for each item was found to 
increase considerably with age. 

The results of Keil's own study (1985, cited in Winner, 1988: 82-
83), which involved children between 5 and 9 years of age, suggest that 
metaphor comprehension 'emerges on a domain-by-domain basis'. Once 
the children understood a metaphor involving items from the two 
domains, such as the properties of plants applied to ideas (e.g. 'The idea 
bloomed'), they also showed understanding of most other metaphors 
involving items from those domains. 

1.4.5 Syntactic Form 

A small number of studies have investigated the influence on com-
prehension of the syntactic form of a metaphor (Nippold, Leonard and 
Kail, 1984; Renolds and Ortony, 1980; Winner, Engel and Gardner, 
1980). The study by Nippold, Leonard and Kail (1984) involved children 
aged 7 and 9 years. Equal numbers of predicative metaphors (involving 
only one topic and one vehicle) and proportional metaphors were 
presented to the children. Proportional metaphors contain two topics and 
two vehicles and involve the use of analogy at an underlying level (e.g. 
'The bird's nest was a piggybank that had no coins' [p. 198]). The 
researchers found that the children had more difficulty understanding the 
proportional metaphors than the predicative ones. The 9-year-olds also 
displayed a greater level of comprehension that than the 7-yearold 
children. 

An earlier study by Winner, Engel and Gardner (1980) (involving 
children aged 6, 7 and 9 years) compared ease of comprehension of five 
different syntactic presentations of metaphor. The researchers had firstly 
hypothesised that the rewriting of a predicate metaphors in the form of a 
simile (a comparison involving the 

30 



 

RIDDLE USE AND COMPREHENSION 

term 'like') would facilitate comprehension, by highlighting that a 
comparison was being used. Secondly, a similar prediction was made in 
relation to topicless metaphors (in which the vehicle is stated but the 
topic is only implied). Topicless metaphors were rewritten both in the 
from of an explicit analogy (e.g. 'A picture frame surrounds a picture 
and ________ surrounds a field' [p. 26]) and in the form of a question 
(which the researchers labelled a 'riddle', e.g. 'What is like a picture 
frame but surrounds a field?' [26]). 

The study found that the original and simile versions of the 
predicative metaphors were equally difficult for the children to 
comprehend. The analogy forms of the topicless metaphors were also no 
more accessible to the children than their original forms. However, the 
'riddle' forms were found to be easier to comprehend than the topicless 
metaphors. The predicative metaphors were also more difficult to 
comprehend than the topicless metaphors. A study by Renolds and 
Ortony (1980), involving a more naturalistic task, similarly reported no 
difference in the ease of comprehension between predicative metaphors 
and similes. Whilst it can therefore be concluded that syntactic form may 
have some influence on the degree of comprehension of metaphor, 
further research is needed to determine the exact nature of this 
relationship. 

1.5 Researcher's Questions 

To date, there is only a small body of research covering the development 
of riddling skills in children. Few studies have documented the 
development of riddle use and it would appear that none of the available 
studies have examined Children's ability to answer riddles. The research 
into riddle comprehension has mainly involved the use of 
grammatically-based joking riddles, as opposed to metaphorically-based, 
'true riddles'. These are all areas, therefore, which are in need of further 
investigation. Most of the studies which have examined children's 
comprehension and use of riddles and metaphors have reported no 
significant difference in the performance of boys and girls in these areas. 
The use of both quantitative and qualitative types of comparison might 
therefore be useful in revealing any subtle differences in the 
performance of the two groups. 

31 



 

1) Are there developmental patterns in the types of answer 
which children give to metaphoric riddle questions? 

STAUNTON 

After reviewing the literature, the following questions were posed 
by the researcher in order to explore the development of metaphoric 
riddling in Irish school-aged children: 

2) Does the ability to comprehend metaphoric riddles (i.e. 
riddle question and answer sequences) improve with age? 

3) Is there a developmental pattern in the types of riddles 
which children use? 

4) Are there differences in the developmental patterns dis-
played by boys and girls in riddle comprehension and use? 

METHODOLOGY 

2.1 Purpose 

The purposes of this study were 

1) to investigate the riddle-answering skills of Irish school-
aged children 

2) to investigate the development of comprehension of 
metaphoric riddles in Irish school-aged children 

3) to investigate the effect of different types of metaphoric 
similarity on comprehension 

4) to investigate the development of children's own use of 
riddles 

5) to investigate the effect of gender differences on the results 
of the study. 

32 



 

RIDDLE USE AND COMPREHENSION 

2.2 Design 

A linguistic approach was used in the research study, incorporating 
elements of an experimental design. 

2.3 Preliminary Procedure 

Selection of Metaphor-Based Riddles 

Twelve riddles which involve the use of metaphor (see Appendix 1) 
were chosen from a wide selection of popular children's riddle and joke 
books. 

TABLE 1: Criteria in Selecting Riddles 

The riddles chosen 

·involved the use of metaphor to provide either ambiguity in 
either the 'question' or 'answer' of the riddle. 

·occurred frequently in the riddle and joke books and were 
therefore more likely to be familiar to the children. 

·included six riddles involving the use of metaphor based on 
sensory grounds and six involving nonsensory grounds. 

·involved a simple 'question and answer' format. 

Construction of Probe Ouestions 

Questions to be used in order to test for comprehension of the twelve 
metaphor-based riddles were constructed. The use of these questions was 
based on the assumption that children who had fully understood a 
metaphoric riddle would be able to state the similarities or 'ground' 
which formed the basis for that metaphor. For example, the metaphoric 
riddle, 'What has legs but cannot walk? (A table) ' is based on a shared 
physical characteristic which has been noticed between a table and a 
human being (or animal). The ground in the case of this metaphor is the 
notion that a table and a 

33 



 

STAUNTON 

human being both have legs. The probe questions which were con-
structed to test for comprehension of the Riddles 1-12 therefore looked 
for the grounds of the metaphors involved and took the form, 'How is X 
like Y?', where the items mentioned were those involved in the 
metaphorical comparison. The task for the children is to verbalise the 
grounds which have already been implied in Riddles 1-12 (presented 
previously). 

2.4 Children Involved in the Study 

Educational Setting 

The children involved in this study attended either:  
(a) an all-female primary school or 
(b) a mixed-sex primary school. 

Both schools are located in an urban setting, are English-speaking 
schools and are judged to be similar in socio-economic status. 

Preliminary Introduction 

A letter of introduction was sent to each school prior to the com-
mencement of the study, along with an explanation of its aims and a 
description of the procedure involved. Discussions were then held with 
the principal teachers, outlining the selection criteria for the children 
involved. A letter explaining the aims of the study and the procedure 
involved was sent to the parents or guardians of children identified as 
suitable for participation in the study. Permission for their children to be 
involved in the study and for audio-taping of the interviews was 
obtained in writing from interested parents and guardians. 

Selection of Children for the Study 

A list of children who were suitable for involvement the study was 
compiled by each class teacher. From this, children who were within the 
average age range for the particular class level (including both boys and 
girls) were identified. Parental permission for possible participation in 
the study was obtained for the children in 

34 



 

RIDDLE USE AND COMPREHENSION 

this group. Children were then chosen at random from the group by the 
researcher for participation in the study. Twenty-four children 
participated in all. Six children were chosen from each of the 3rd, 4th, 5th, 
and 6th classes. Equal numbers of boys and girls were chosen from each 
of the class levels. 

TABLE 2: Selection Criteria 

The children involved in the study 
 ·were within the average age range for their class levels. 

·were judged by their teachers to be developing normally and 
to be of average ability. 

·had no reported speech and language difficulties or difficulties 
with written language. 

·had the written permission of their parents or guardians to take 
part in the study. 

TABLE 3: Class Levels and Average Age Ranges 

CLASS 

3rd Class  
4th Class  
5th Class 
6th Class 

AGE RANGE FOR AVERAGE AGE RANGE 
 THE CLASS LEVEL 
 8yrs 2m to 8yrs 8m 
 9yrs 3m to 9yrs 9m 
 10yrs 1 m to 10yrs 7m 
 11 yrs 0m to 11yrs 6m 

8yrs 0m to 9yrs 4m  
8yrs 11m to 10yrs 2m  
9yrs 10m to 10yrs 11m 
10yrs 10m to 12 yrs 8m 

TABLE 4: Details of the Children Selected 

CLASS 
3rd Class  
4th Class  
5th Class  
6th Class 

CHILD NO.  
1-6 

7-12 
13-18 
19-24 

AVERAGE AGE  
8 years 5 months  
9 years 5 months 
10 years 3 months 
 11 years 4months 

35 



 

STAUNTON 

2.5 Procedure 

The procedure was informal in nature and essentially took the form of a 
riddling session between the researcher and each child. There were three 
main tasks involved in the procedure, however the order of presentation 
of these tasks was frequently altered depending on the interests of the 
children involved. For example, if a child was especially eager to begin 
telling riddles at the start of the interview, then the third task was 
completed first. 

The children were interviewed individually by the researcher. Each 
interview was audio-recorded for later analysis. The interview began 
with the researcher introducing herself to the child and explaining that 
she was doing a project which was looking at the riddles which children 
use. The riddle tasks were then introduced with the following 
explanation: 

'I'm going to tell you some riddles to see if you've heard any of them 
before and then afterwards you can tell me any riddles that you know. 
Some riddles that you might have heard before are ones like, 'Why did 
the chicken cross the road? (To get to the other side) 'or 'What goes up 
when the rain comes down? (An umbrella)' or 'What has hands but no 
arms? (A clock) '. 

The child was given a chance to answer the introductory riddles 
(above) if he or she was able. Otherwise, the answers to the riddles were 
given by the researcher. Further examples of riddles were given if 
necessary, until the child understood the 'question and answer' nature of 
the riddles. Three tasks were then presented to the child: 

Task 1: Listening to Riddles 

The first task (Task I) involved the presentation of the metaphoric 
riddles. The task for the child was to listen to Riddles 1-12 (see 
Appendix 1) and their riddle answers. The children could also attempt to 
answer the riddles if they so wished The presentation of each riddle-
question was followed by a pause, during which the child had an 
opportunity to answer the riddle or to respond, 'I don't know'. Children 
who attempted to answer the riddle were encouraged to state why they 
had given that answer. The researcher 

36 



 

RIDDLE USE AND COMPREHENSION 

then responded, 'That’s very good. Another answer that I was going to 
say was. . . " and provided the child with the target response. Similarly, 
children who did not attempt to answer the riddle were provided with 
the target response. 

Task 2: Responding to Probes 

This task was introduced with the instruction, 'I'm going to give you 
some more questions now which are a bit the last ones. You listen to 
them and try to think of the answers '. The questions which had been 
constructed to probe the children's comprehension of the riddles were 
then presented in random order (i.e. in a different order to their 
corresponding riddles). The children were encouraged to listen to each 
question and try to think of the answer. Some of the terms which were 
involved in the nonsensory Riddles (7-12) were in more abstract than 
those involved in the sensory Riddles (1-6). The depth of the children's 
comprehension of the metaphors involved in Riddles 7-12 was therefore 
probed further by encouraging the children who gave correct answers to 
elaborate further. Prompts such as, 'Can you tell me a little bit more 
about that? 'or, 'What is a bright person like?', were therefore used to 
probe for full understanding of the non sensory metaphors. 

Task 3: Riddle-Telling 

Finally, the children were asked, 'Do you know any riddles a bit like the 
ones that I've told you? ' and the riddles which they told were recorded 
Riddles which children contributed at any other point during the 
interview were also recorded for later analysis. 

2.6 Analysis Protocol 

The interviews were later transcribed from the audiotape. The responses 
collected from the children on each of the three tasks were analysed both 
quantitatively and qualitatively, where applicable. Patterns which 
emerged in the style of answering on all of the tasks were recorded and 
described. 

37 



 

STAUNTON 

RESULTS 

3.1 Introduction 

This study sought to investigate the development of children's ability to 
both comprehend and pose metaphoric riddles. The following questions 
in particular were investigated: 

1) Are there developmental patterns in the types of answer 
which children give to metaphoric riddle questions? 

2) Does the ability to comprehend metaphoric riddles increase 
with age? 

3) Is there a developmental sequence in the types of riddles 
which children use? 

4) Are there differences in the developmental patterns dis-
played by boys and girls in riddle comprehension and use? 

3.2 Development of the Ability to Answer Metaphoric Riddles 

Although the main aim of Task I (where Riddles 1-12 were presented) 
was simply to allow the children to listen to the riddles, the informal 
nature of the interview meant that many children attempted to answer the 
riddles also. The task involved in answering the riddles which were 
based on nonsensory grounds (7-12) was much different to that involved 
in answering the sensory riddles (1-6). In compiling the list of riddles 
prior to testing (from riddle and joke books), it became apparent that, in 
general, riddles based on nonsensory grounds only contain the use of 
metaphorical language to provide ambiguity in their answers, rather than 
in the riddle questions. In order for the children to answer riddles 7-12, 
they were therefore required to have made use of metaphorical language 
in their answers. From the sample of 216 answers to these non sensory 
riddles, only two answers were judged as possibly completing the 
riddles. A large proportion of the other answers were 'Don't know' 
responses. Analysis of the children's answers therefore focused on their 
responses to the riddles (1-6) involv- 

38 



 

RIDDLE USE AND COMPREHENSION 

ing sensory metaphors, which are more typical of 'true' riddles. 
The answers given by the children to the true riddles were analysed 

by the tester and classified in three ways. The answers were firstly 
considered to be 'target' or canonical answers if they were the same as, or 
very similar to the answers which had been found in the riddle and joke 
books. Secondly, children in all of the age groups gave a number of 
'alternative' answers to the riddles, which had not been found in any of 
the riddle or joke books. All of the alternative answers given by the 
children were judged as fitting the descriptions given in the riddles to 
some degree, although some were deemed more suitable than others. 
This finding will be described in more detail below. The final group of 
answers were 'Don't know' responses. 

An examination of the results firstly from a quantitative point of 
view revealed some possible developmental patterns. The overall 
accuracy of the children in supplying the target answers to the riddles 
was found to be greater amongst the older children as compared with the 
younger groups. However, the 8-year-old group were found to have 
given more target answers than the 9-yearolds. Whilst the 9-year-old 
group did not give the target answer for any of the riddles, they were the 
group most likely to give an 'alternative' type of answer to the riddles. 
The proportion of target answers given increased greatly after 9 years, to 
just over a quarter of the answers of the 10-years-olds and two-fifths of 
the answers given by the 11-year-olds. 

About half of the answers given by the 8- and 9-year-olds were 
classified as 'alternative' answers. The incidence of alternative-type 
answers decreased overall amongst the two older groups, with just under 
a quarter of the answers of the 10-year-olds being of this type. The level 
was slightly higher amongst the 11-year-olds, at about one-third of 
answers. The proportion of 'Don't know' responses gradually increased, 
from just over two-fifths of answers at 8 years to half of the answers at 
10-years-of-age. The proportion of 'Don't know' answers was lowest 
amongst the 11-year-olds (the oldest group), and this was also the group 
who were most likely to respond with the target answer. 

39 



 

STAUNTON 

Analysis of the Alternative Answers 

A second type of analysis was also used to examine the 'alternative' 
riddle answers generated by the children. This involved the examination 
of the answers in terms of their ability to satisfy the condition known as a 
'partially obscured semantic fit' which exists between the riddle question 
or 'image' and its answer (Scott, 1965, cited in Kallen, 1981: 4). In order 
to fulfill this condition, the riddle answer firstly had to belong to a 
semantic field opposite to that of the riddle image. 

The second condition which the children's answers had to fulfill was 
that of being judged a prototypical example of a member of the particular 
semantic category. The final criteria which was used was that the riddle 
answer had to share some obvious, minor characteristic with the item in 
the riddle image in order to complete the metaphorical comparison. 

The first kinds of answers which the children gave to the riddles 
were considered to be 'anomalies' because they did not strictly fit the 
description given in the riddle image, even in literal terms. There were 
just a couple of instances of such answers, both of which were given by 
9-year-old children. These answers belonged to the same superordinate 
category as the item implied in the riddle image (i.e. animate) and were 
therefore not considered to have a great degree of obscurity as answers. 
However, in the case of each of the responses of this type, the children 
had created an explanation to make the answer fit the description in the 
riddle image, as can be observed from this example: 

Image: What can howl but has no mouth? Answer and Explanation: 'A 
frog or a toad, because a toad can let out a big noise and their mouths 
are different to ours '. 

Just over a fifth of the answers given by the children were those 
which were considered to be 'literal' answers as they fitted the 
description used in the image in a literal sense, and there was no attempt 
to give an answer which had originally been obscured through the use of 
metaphor. The riddle answer was therefore from the same category 
(animate) as the riddle image in each case. This type of answer was 
given with the greatest frequency amongst the 8- and 9-year-olds, with 
few instances amongst the older children. Examples include the 
following: 

40 



 

RIDDLE USE AND COMPREHENSION 

Image: What has legs but cannot walk? Answer: Tadpoles.  
Image: What has eyes but cannot see? Answer: A bat.  
Image: What goes around all day with its tongue sticking out and sits 
by the bed at night? Answers: A dog, A cat. 

Answers of a similar type to those above involved the use of a feature 
named in the riddle image, put in a situation so that it fitted the 
description in the image as in this example: 

Image: What has teeth but no mouth? Answers: Teeth in a box. 

The third group of answers were those which came from the same 
superordinate category or set as was implied in the riddle image (i.e. 
animate), but involved some change of state so that the human or animal 
was made to fit the description in the image, as in these answers: 

Image: What has legs but cannot walk? Answers: A person that has a 
spell put on them. 
Image: What has eyes but cannot see? Answers: A dead fish. 

Just over one-tenth of the answers in the study were of this nature 
and were given by children from all age groups. Examples of the changes 
of state described include the use of magic, the death of the animate 
entity or the existence of a disabling condition. 

The remaining types of answers given in the study were those which 
were considered to have some degree of validity as conventional riddle 
answers, as they came from an opposite category to the item implied in 
the image (inanimate) and were obscured to some extent. However, the 
requirement of a 'partially obscured semantic fit' was judged to be 
fulfilled more readily by some of these types of answers than others. 
Some of the answers given were therefore judged as being more suitable 
in terms of completing the riddle than others. The fourth overall group of 
answers were labeled 'replica-type answers' due to the fact that they in-
volved the naming of an object which was considered to be a rep-
resentation of a living thing. These answers were therefore considered to 
be partially obscured in that they came from the opposite superordinate 
category to the human or animal implied in the 

41 



 

STAUNTON 

riddle (i.e. they were inanimate objects). They were also man-made or 
'cultural' objects, as opposed to 'natural' objects, according to Köngäs-
Maranda's (1974) method of classification. However, they were not 
considered prototypical examples of inanimate objects as they were 
generally objects which bore a likeness to a real living creature or part of 
a living creature (e.g. a teddy bear can be thought of as an inanimate 
representation of a real bear). These answers accounted for about one-
sixth of the responses given by the children and occurred with greatest 
frequency amongst the 9and 10-year-olds and included the following: 

Image: What has legs but cannot walk? Answer: A doll or a teddy; A 
gymnastic horse (i.e. a wooden horse used in gymnastics). 
Image: What has eyes but cannot see? Answer: Joke glasses (with 
eyes drawn on them) Cats' eyes on the road 

A similar group of answers were those which, although they were 
not considered to represent a living thing, were considered to be linked 
very closely to the animate item implied in the riddle image. These 
answers came from the inanimate category but were not considered to be 
obscured to a degree sufficient to provide good metaphor in the riddle. 
For example, the answer, 'trousers', given for the riddle image, 'What 
has legs but cannot walk? " was considered to be partially obscured 
because the item is inanimate, but was judged to be too closely related to 
a human leg, as a trouser leg actually fits over a human leg. It was 
therefore not considered a good example of a metaphorical renaming. 
Similarly, the answer given by two children to the riddle image, 'What 
runs but never walks?' was 'a runner' (running shoe). Such an answer 
does not involve a metaphorical renaming, and instead involves the use 
of a name closely related to the action named in the riddle image. 
Overall, there were only a few answers in this fifth group and they were 
given only by the 8- and 9-year-old children in the study. 

The sixth and largest group of answers in the study (accounting for 
just over two-fifths of responses given) were those which were 
considered to have most validity as answers for each of the riddles. They 
involved the naming of items which belonged to the inanimate set and 
which could also be considered to involve good examples of metaphor. 
These answers involved the use of meta- 

42 



 

RIDDLE USE AND COMPREHENSION 

phor of two kinds. Firstly, the 8- and 9-year-olds gave answers which 
were characteristic of 'radical polysemy'. Such answers involved the 
creative metaphorical renamings of an object or an action by the 
children, based on some type of similarity which they had noticed. This 
is illustrated in the following examples: 

Image: What has teeth but no mouth? Answers: A piano; A needle 
(one tooth). 
Image: What goes around all day with its tongue sticking out and sits 
by your bed at night? Answers: (The safety sticker on) a lamp; Your 
tie. 

The remaining answers involved the use of 'conventional polysemy' 
as they were metaphorical terms which are already present in the 
language (e.g. the 'eye' of a needle, the 'running' of an engine). Almost a 
third of the answers given in the study were of this type and their use 
was most common amongst the wand 11-year-olds. The two meanings 
of terms involving conventional polysemy are intuitively thought of by 
native speakers as being related. It was this type of metaphor which was 
in use in the original 'target' answers for the riddles, such as the 'leg' of a 
table or the 'howling' of the wind. These answers therefore tended also to 
involve more prototypical examples of inanimate objects than the other 
types of answers which were given (such as the 'repilca-type answers'). 
These answers included the following: 

Image: What has teeth but no mouth? Answers: A trap; A saw.  
Image: What has eyes but cannot see? Answers: A needle; A pin 
(meaning 'a needle '). 
Image: What runs but never walks? Answers: Your nose; A car 
engine; A tape recorder. 

The children as a group were judged to be more successful in 
providing answers which fulfilled the condition of a 'partially obscured 
semantic fit' for the first four of the six riddles based on sensory 
similarity. The contrast involved in the target answers for Riddles 1-4 
was that between an animate entity and an inanimate one, but more 
specifically between a human being and a cultural object. This contrast 
was made in a number of the answers which 

43 



 

STAUNTON 

were suggested by the children. However, the contrast involved in 
Riddles 5 and 6 (when the original answers are used) was that between 
an animate entity (human or animal) and a natural object (i.e. water, the 
wind). The answers which the children suggested, although valid in that 
they were from the inanimate set, an belonged to the set of cultural 
objects, such as the following answers: 

Image: What runs but never walks? Answers: A car engine; A tape 
recorder. 
Image: What can howl but has no mouth? Answers: A radio; The 
alarm system (burglar alarm); A hairdryer; An alarm clock; The 
brakes on a car. 

The set of natural objects did not feature in any of the answers given by 
the children in the study. 

Summary: Ability to Answer True Riddles 

Overall, the children's accuracy in giving 'target' or canonical answers 
increased with age. The only exception to this pattern was amongst the 
9-year-old children who gave less target answers than the 8-year-olds. 
Children aged 8 and 9 years were also the most likely to give a response 
other than the target answer. The 9-yearolds also gave the most 
alternative answers of any of the groups. The 11-year-olds were the 
group least likely to give a 'don't know' response and most likely to give 
the target response. 

The riddle answers classified as 'alternative' answers were analysed 
in terms of their ability to complete the condition known as a 'partially 
obscured semantic fit' between the riddle image and the answer. Overall, 
the 10- and 11-year-olds, were more likely to give an answer which was 
from a category opposite to that of the item implied in the riddle image. 
The prototypicality of the riddle answer which belonged to an opposite 
category was also greatest amongst the 10- and 11-year-olds. 'Literal' 
types of answers were characteristic of children aged 8 years, while 
'replica-type' answers occured with greatest frequency amongst the 9- 
and 10-year-olds. The answers judged as metaphorical which were given 
by the 8- and 9-year-olds tended to be more novel than those of the 10- 
and 11-year-olds, who made use of metaphors already sanctioned in the 
language. 

44 



 

RIDDLE USE AND COMPREHENSION 

The patterns which emerged overall in the types of answer given by 
the children at different age levels are summarised in Table 5 below. 

TABLE 5: The types of answers given by the children in each age group 
to the riddles based on sensory similarity (n = 24). 

 8 years 9 years 10 years 11 years 

Target 
Answers 11% 0% 28% 39% 

Alternative 47% 53% 22% 30.5% 

'Don't Know' 
Responses   44% 47% 50% 30.5% 

3.3 Development of the Ability to Comprehend Metaphoric Riddles 

The method used to test for comprehension of metaphoric riddles (i.e. 
riddle question and answer sequences) consisted of two tasks. The 
children were firstly posed twelve metaphoric riddle questions and given 
their answers (Task 1). They were later asked randomly ordered 
questions based on the riddles which looked for the ground of each 
metaphor (i.e. the children had to state the type of similarity mentioned 
in the riddles earlier). This was Task 2. The 'probe' questions were of the 
form, 'How is X like V?' or 'How could X be like Y?', where the items 
mentioned were the topic and vehicle involved in the metaphorical 
comparison. As stated earlier, the use of these questions was based on 
the assumption that children who had fully understood a metaphoric 
riddle would later be able to recall the similarities or ground which 
formed the basis for that metaphor. Children who were able to state the 
grounds of the metaphoric riddles were therefore judged to have 
understood these riddles. 

In examining the results of this task, some developmental patterns 
emerged. Overall, the number of children judged as being able to recall 
the grounds of the metaphoric riddles on Task 2 was seen to increase 
with age. This therefore suggests increased comprehension of metaphors 
in the context of riddles amongst the older children as compared with the 
younger ones. The answers which 

45 



 

STAUNTON 

the children gave on this task were analysed as either being correct (i. e. 
the target response), an alternative response, or a 'don't know' response. 

A comparison of the results indicates that the greatest difference in 
performance between children in consecutive age groups is between the 
8- and 9-year-olds. The 8-year-olds were able to recall the grounds of 
the riddles with 72% accuracy compared with the 80% level of accuracy 
achieved by the 9-year-olds. However, the 75% level of accuracy 
achieved by the 10-year-olds was lower than that of the 9-year-old 
children. The 11-year-olds displayed the greatest level of 
comprehension and were able to recall the grounds of 83% of the 
riddles. 

The number of' don't know' responses was found to decrease overall 
amongst the older children. Whilst just under a fifth of the 8-year-olds 
gave a 'Don't know' response, only one-tenth of the 11-year-olds gave 
such a response. However, the number of 'alternative' responses given 
was greatest amongst the 9- and 10-year-aids. This finding is described 
in more detail in section 3.3.2. 

Summary: Comprehension of Riddles Involving Metaphor 

The ability to comprehend the metaphoric similarity involved in the 
riddles presented was observed overall to increase with age, as judged 
through the use of a 'probe' task. Only the performance of the 10-year-
olds was not in keeping with this pattern. The greatest increase in the 
ability to comprehend the riddles appeared to occur between the ages of 
8 and 9 years. The 11-year-olds achieved the greatest level of accuracy 
overall in recalling the grounds of the metaphoric riddles. A quantitative 
summary of the overall patterns of responding on the task probing for 
comprehension is given in Table 6 below. 

TABLE 6: Types of response given by children in different age 
groups on the task testing for comprehension of the metaphoric 
riddles (n = 24). 

Expected Explanation 
Alternative Answer 
'Don't Know' Response 

8 years 
72 % 
10% 
18% 

9 years 
80% 
17%  
3% 

10 Years 11 Years 
75%  83  
12.5%  8.3% 
12.5%  8.3% 

46 



 

RIDDLE USE AND COMPREHENSION 

3.3.1 The Effect of Type of Similarity on the Comprehension of 
Metaphoric Riddles 

The riddles presented to the children were of two types (see Appendix 1). 
The first six were metaphoric riddles which demanded that the listener 
view the topic of each metaphor in terms of its vehicle, based on shared 
sensory grounds. The topic and vehicle are linked in these sensory 
metaphors due to similarities of appearance (e.g. the physical appearance 
of the leg of a table likened to that of the leg of a person), of sound (an 
animal's howling likened to the howling of the wind), or of movement 
(the movement of a person likened to the movement of water). The 
remaining riddles (7-12) also involved metaphorical uses of language, 
though the similarities between the topic and vehicle were of a 
nonsensory nature. The nonsensory metaphors were of either a 
psychological-physical type (Riddles 7 and 8) or a relational type 
(Riddles 9-12). 

The most striking pattern when comparing the overall performance 
on the task which probed for comprehension of the riddles (Task 2) was 
the difference in the children's ability as a group to comprehend sensory 
as opposed to nonsensory riddles. Examining the results firstly from a 
quantitative point of view, the children were shown to comprehend the 
sensory riddles (1-6) with between 75% and 88% accuracy overall (as 
judged by the use of the probe question task). However, their level of 
comprehension of the nonsensory metaphors involved in Riddles 7-12 
was lower, ranging between 59%-78% overall, as judged by the same 
task. 

A comparison of the performance of each age group on the probe 
question task indicates firstly that there were no large differences 
amongst the age groups in their level of comprehension of the riddles 
involving sensory metaphor. The 8- and 9-year-olds were judged to have 
comprehended the sensory metaphors with an 86% level of accuracy, and 
this increased very slightly for the 11-year-olds. The performance of the 
10-year-olds on this task (who comprehended 75% of the sensory 
metaphoric riddles), however, was not in keeping with the performance 
of the other age groups. Secondly, the correct answers which were given 
on the probe question task for the sensory riddles were found to be quali-
tatively very similar across the age groups. 

47 



 

STAUNTON 

A more noticeable pattern emerged in the abilities of the children to 
verbalise the similarities involved in the nonsensory metaphors. A 
comparison of the results from the younger children with those of the 
older group reveals both quantitative and qualitative differences on this 
task. The 8-year-olds verbalised the similarities involved in the 
nonsensory metaphors with a 59% level of accuracy (as measured by the 
use of the probes on Task 2). This level increased considerably to 75% 
for both the 9- and 10-yearold children, and was slightly higher for the 
11-year-olds, at 78%. 

A further pattern emerged in the depth of comprehension of the 
nonsensory riddles which was indicated by the responses given by the 
children. Many of the younger children (8- and 9-year-olds) correctly 
recalled the similarity which was earlier implied in each riddle and were 
therefore deemed to have understood the metaphor involved. In the case 
of Riddle 7, for example, the question which probed for comprehension 
of the nonsensory metaphor was, 'How could a person be a bit like the 
sun?', to which the accepted answer was 'If they're bright' or 'They could 
both be bright'. Although it was assumed once they had given the correct 
answer that the younger children were aware that the terms could be ap-
plied in two senses, their explanations (when prompted further) were 
generally less accurate and less complete than those of the older group 
(10- and 11-year-olds). 

In the case of the riddles involving psychological-physical 
metaphors (which involved the dual-function adjectives 'bright' and 
'hard'), the 8- and 9-year-olds tended to verbalise only the physical sense 
of the terms in their explanations, as opposed to both senses. Examples of 
such explanations include the following: 

Question: How could a person be a bit like the sun? Child (8 years of 
age): 'If they were bright '. (Further prompt): '/f someone's bright, their 
teeth would be all shiny'. 

Question: How could a person be a bit like concrete? Child (9 years of 
age): 'Because they could be hard as well'. (Further prompt): 'Like, 
cement would be really hard and a hard person can hurt if they bump 
into you '. 

The explanations given by the 10-year-olds more frequently 
contained reference to both the psychological and physical mean 

48 



 

RIDDLE USE AND COMPREHENSION 

ings involved in the metaphors than those of the younger children. Only 
the 11-year-olds who had given the target answer were consistently able 
to articulate the fact that the terms could be used to refer to both an 
object (in the literal sense) and a person (in the psychological sense). 
Responses which were more typical of the older children include the 
following: 

Question: How could a person be a bit like the sun? Child (10 years of 
age): 'They could be bright '. (Further prompt): 'Well, the sun could 
be bright because it shines a lot, but a bright person would be really 
clever and good at their work. And they can kind of shine like that 
because they're so clever…' 
Question: How could a person be a bit like concrete? Child (11 years 
of age): 'Both of them could be sort of hard'. (Further prompt): 
'Something might happen to them and they would take their anger out 
on other people and stuff. 

The ability to verbalise both senses of the relational metaphors 
involved in Riddles 9-12 appeared overall to be present slightly earlier in 
this group of children than the ability to explain the psychological-
physical metaphors. The ability to articulate the relationship between 
both meanings of the relational metaphors was indicated in some of the 
responses given by the 8- and 9-year-old children. For example, the 
following responses were given by children in this age group: 

Question: How can time be a bit like a bird? Child (8 years of age): 'It 
can fly and a bird can fly'. (Prompt): 'When time flies it means that it 
goes really fast '. 

Question: How could a person be a bit like a broken up jigsaw? Child 
(9 years of age): 'A broken up jigsaw falls to pieces and the person 
could fall to pieces because they're sort of sad and want to do 
something but they can't do it '. 

Children in all the age groups who were unable to verbalise the 
similarity involved in the relational metaphors tended to give a 'Don't 
know' response, rather than an incomplete or inaccurate response. By 10- 
and 11-years-of-age, the children who gave correct responses to the 
questions probing for comprehension of the 

49 



 

STAUNTON 

relational metaphors were consistently able to articulate the grounds of 
these metaphors. The following examples are typical of such 
explanations: 

Question: How could a telephone be a bit like a person who’s going to 
get married soon? Child (10 years of age): 'It could be engaged '. 
(Prompt): 'If a telephone is engaged it means that it s busy because 
someone is talking on it when you're trying to ring them '. (Prompt): 
'The person who s engaged is busy because they're going to get 
married to someone soon '. 

Question: How could a person be a bit like a sweeping brush? Child 
(11 years of age): 'Because they could be pushed around'. (Prompt): 'If 
someone gets pushed around it means that people don't respect them or 
take any notice of them. And people make a sweeping brush clean up 
by pushing it around the floor '. 

Summary: The Influence of the Type of Similarity on Com-
prehension 

Riddles involving a sensory type of metaphoric similarity were found to 
be more easily comprehended by children from every age group than 
those involving nonsensory similarity. The level of comprehension of the 
sensory riddles was found to vary little with age. The 10-year-olds, 
however, achieved a lower level of accuracy than the younger children 
for this type of riddle. 

Comprehension of riddles involving a nonsensory type of similarity 
increased greatly between 8 and 9 years, but varied little beyond that 
age. The older children in the study (10- and 11-year-olds) indicated a 
greater depth of comprehension of the nonsensory metaphors, when 
elaborating on their responses to the probes, than the younger groups. 
Overall, the ability to verbalise both meanings of the relational type of 
metaphor was observed to emerge earlier (at 8 or 9 years) than in the 
case of the psychological-physical type of comparison (at 10 years). The 
accuracy of each age group in answering the probe questions for the 
sensory and nonsensory riddles is summarised in Table 7 below. 

50 



 

RIDDLE USE AND COMPREHENSION 

TABLE 7: Accuracy of each age group in comprehending riddles 
involving sensory and nonsensory metaphors (as judged by the use of 
the probe question task) (n = 24). 

 Sensory Metaphors Nonsensory Metaphors 

8 years 86% 59% 

9 years 86% 75% 

10 years 75% 75% 

11 years 88% 78% 

3.3.2 The Ability of the Children to Remark on Types of Similarity 
not Implied in the Riddles 

The ability to remark on a similarity between two entities might 
essentially be said to be the ability which is involved in creating a 
metaphor. The children in the oldest age group tested in this study (the 
11-year-olds) were the most likely to verbalise the similarity involved in 
the riddles which they had heard earlier (for Task 2), which was the 
desired response on this task. However, in analysing the results it also 
became apparent that a number of children who were unable to correctly 
recall this similarity were able to compare the items involved on other 
grounds which were not implied in the original riddles. A number of 
these responses were highly thoughtful and creative. Additionally, some 
of the children who were able to recall the similarity implied earlier in 
the riddles added to their answers by giving other ways in which the 
items involved are similar. 

The incidence of the 'alternative' responses was greatest amongst the 
9- and 10-year-olds in the study. Examples of these responses are given 
below. A further pattern emerged in the type of responses collected from 
the children from different age groups. The younger children in the 
group (8 years of age) were most likely to give creative responses which 
were based on a sensory similarity between the two items. The sensory 
similarities which were noted were either physical in nature or were 
similarities of movement. Responses which were typical of the 8-year-
olds include the following: 

51 



 

STAUNTON 

Question: How is a table a bit like a person? Child (8 years of age): 
Because they have legs. And the table stands still when the TVs on and 
so does a person. 

Question: How is a potato a bit like a face? Child (8 years of age): 
'Because sometimes it has dots somewhere. And supposing the dots got 
stuck together and another dot went up like a mouth, it would look like 
a face '. 

The creative responses of children aged between 9 and 11 years were 
more frequently based on a relational type of similarity than any other 
type of similarity. Such responses remarked upon a similarity between 
two items in the ways in which they function (as opposed to a purely 
physical type of similarity). Although the boundary separating relational 
and psychological-physical metaphors is often vague (Winner, 1988), 
there appeared to be few examples of psychological-physical metaphors 
amongst the creative responses. Examples of the responses grounded in a 
relational type of similarity include the following: 

Question: How is a shoe a bit like a mouth? Child (9 years of 
age): 'Because you put your foot into it. It's like the shoe is kind of 
eating your foot. But If you take your foot out, it's not kind of slobbery 
or anything. And you can 'put your foot in it' if you 're talking and you 
say something you shouldn't say'. 

Question: How could a person be a bit like the sun? Child (10 years of 
age): 'If it 's bright and it gives off rays. You know, like if people are 
clever they kind of give their information out. And the sun sends 
affrays. A clever person gives rays of information out to other people, 
you know?' 

Question: How could a person be a bit like a rock? Child (11 years 
old): 'If they 're tough '. (Prompt): A tough person might have their 
mind made up about something and it would be really hard to 'get 
around' them, to get them to change their mind. And if there's a really 
big rock in the middle of the road or somewhere then it would be really 
hard to get around that too '. 

In analysing the responses it also became clear that individual 
children were more likely than others to give creative responses. 

52 



 

RIDDLE USE AND COMPREHENSION 

For example, the following responses were all given by a ten year-old 
girl: 

Question: How is a table a bit like a person? Child. 'They have legs 
and they make you eat. As soon as you see the table you kind of go, 
'Grand, this means food. Just sit down, eat!' And I don't like vegetables 
or potatoes or anything like that and my Mum makes me eat them '. 

Question: How is a shoe a bit like a mouth? Child: 'It has a tongue. 
And if it's old, it opens too much. You know the way if the shoe is old, 
the top of it comes apart and it opens. And when you get older you tend 
to talk a lot more, so your mouth opens more. Like, my granny used to 
talk a lot... '. 

Question: How can the wind be a bit like a wolf? Child: 'It howls. And 
it can carry things, like frogs or fish. Like sometimes you can have 
rain with a frog in it because of the way the water evaporates up into 
the cloud and the frogs could be taken up into the clouds with the 
 fish '. 

Summary: Ability to Comment on Similarities not Implied in the 
Riddles 

Children who were unable to recall the target type of similarity in the 
riddles presented earlier were sometimes able to compare the topic and 
vehicle based on other grounds. The incidence of these 'alternative' 
answers was greatest amongst the 9- and 10-yearolds. The creative 
explanations given by the 8-year-olds were all based on a sensory type of 
similarity. Children between the ages of 9 and 11 years were more likely 
to base their creative explanations on a relational type of nonsensory 
similarity. A couple of instances of explanations based on psychological-
physical grounds were also recorded amongst the older children. Overall, 
some children were more likely than others to make use of a creative 
type of response rather than a 'Don't Know' response. 

3.4 Development of the Use of Riddles 

The final task in the study involved the children being asked to tell the 
researcher any riddles which they knew. The riddles which 

53 



 

STAUNTON 

were collected were analysed in terms of both their overall structure and 
the types of 'block element' involved, according to the methods of 
classification described by McDowell (1979) in his study and by Green 
and Pepicello (1984). Having been asked to supply the researcher with 
some riddles, many children supplied a number of jokes also, perhaps 
viewing the riddles and jokes as being linked through their common 
property of humour (see Appendix 3). In collecting and analysing the 
sample of riddles it also became obvious, however, that the boundaries 
between what constitutes a riddle and a joke are not always well defined. 
For example, some of the responses which were classified by the re-
searcher as 'jokes' made use of the device known as 'catching' which is a 
common feature of riddles. It must therefore be borne in mind when 
examining the results that the division between the concepts of 'joke' and 
'riddle' is often an arbitrary one. Whilst the jokes which the children 
supplied were not analysed in detail, a cursory examination indicates 
that metaphor was not used as a source of ambiguity in any of the jokes 
in this sample. 

Analysis of Riddle Structure 

The riddles which were collected from the children in this study were 
firstly analysed in terms of their overall structure (according to the types 
of structure described by McDowell, 1979: 66-68). Three main types of 
structure featured in the present sample. 

Over half the riddles and descriptive routines in this sample were 
based on attribution, which involves the use of many simple descriptive 
units, which are described as being 'stacked'. Riddles of this type were 
most common amongst the 8-year-olds (though half of these were 
descriptive routines) and the 11-year-olds. Examples of these riddles 
include the following: 

What’s got two legs, a medium-sized body and a very, very big tail? A 
peacock. (Girl, aged 8 years) 

What do you call a chimp with a banana in each ear? Anything you 
like—he can't hear you! (Boy. aged 10 years) 

What do you call a cucumber with four doors? A slammer. (Girl,  
aged 10 years) 

54 



 

RIDDLE USE AND COMPREHENSION 

Almost a third of the riddles in this study were concerned with 
causality and usually took the form of a question involving 'Why' and 
answer beginning with 'Because'. This type of structure was used most 
by the 8- and 9-year-olds. Examples of this nature included the 
following: 

Why did the egg go around the world? Because he wanted to be an 
eggs-plorer. (Boy, aged 10 years). 

Why did the boy bury his clock? Because the batteries were dead. 
(Girl, aged 12 years). 

Riddles based on instrumentality did not feature in the sample 
collected in this current study. The two other riddle structures which 
McDowell (1979) describes, take the form, 'What did the X say to the Y?' 
and 'What kind of X is a P'. Only the former construction featured in the 
present sample of riddles, whereby inanimate objects appear to be 
engaged in imaginary conversations and the two possible meanings of a 
well-known figure of speech are exploited. This structure was used in 
approximately a quarter of the riddles in the current study and its use 
decreased steadily amongst the older children. Typical examples include 
the following: 

What did the big biscuit say to the little biscuit? 'Crumbs '. (Girl,  
aged 8 years). 

What did the skeleton say to the toilet? 'You can't get anything out  
of me '. (Boy, aged 8 years). 

Analysis of Types of Riddling Device Used 

The content of the riddles was analysed according to the types of block 
element or other riddling device which were used to provide ambiguity 
or confusion, using the descriptions of riddling devices given by Green 
and Pepicello (1984) and McDowell (1979). The riddles were each 
assigned to what was judged to be the most appropriate category in each 
case, although in reality some of the riddles could be classified as 
belonging to more than one category. 

55 



 

STAUNTON 

There were firstly some examples of descriptive routines given by 
an 8-year-old girl which, by definition, do not contain a block element 
and involve the description of one or more attributes of an object. The 
routines which this child supplied were as follows: 

What's big and has a very big appetite? An elephant. 

What can you get that can be different colours and very soft? A 
cushion. 

What’s got two legs, a medium-sized body and a very, very big tail? A 
peacock. 

What’s green and brown and has different colours all over it? A 
Christmas tree. 

The rest of the riddles can be classified as 'riddles proper' as they all 
employed some type of device which was intended to cause confusion on 
the part of the riddlee. Riddles involving block elements of both a 
grammatical and a metaphorical variety were collected in the study. Of 
the grammatically-based riddles collected from the children in this study, 
only riddles which made use of phonological and morphological types of 
ambiguity were recorded. Phonological ambiguity was the device used in 
a quarter of the riddles collected in the study and simple lexical 
ambiguity was the device used more often than any other by the children. 
Included here were words which involved homophony in that they made 
use of a single phonetic string which had two unrelated meanings. 
Riddles such as, 'What did the big chimney say to the little chimney? 
('You're too young to be smoking) ' or 'Why did the tortoise cross the 
road? (He wanted to get to the Shell station) ' made use of two possible 
interpretations of a single word. The use of this type of ambiguity 
increased up to 10 years, but decreased sharply amongst the 11-year-
olds. The second type of single-word homophony was that of 
conventional polysemy, in which the two meanings of the word are 
thought of as being related. This will be described along with the riddles 
involving the use of metaphor. Also at the phonological level, there was 
just one example (given by a 8-year-old girl) of the use of ambiguity 
involving minimal pairs ('paws' and 'Claus'), in the riddle: 'Why does a 
lion like 

56 



 

RIDDLE USE AND COMPREHENSION 

Christmas? Because of his sandy paws (Santa Claus) '. No other types of 
phonological ambiguity were featured in this sample of riddles. 

Ambiguity at the morphological level accounted for about one-tenth 
of the riddles collected and was only used by the 9- and 10year-olds in 
the study. The riddles involving morphological ambiguity were of three 
kinds. Firstly, the riddle given by a 10-yearold girl, 'What’s black  
and white and read/red all over? (A newspaper) ' involves the use of 
homophony between the past participle morpheme of the verb 'to read' 
and the colour adjective 'red'. Secondly, the riddle 'Why did the egg go all 
around the world? (Because he wanted to be an eggs-plorer) ' involves 
the deliberate confusion of the word 'egg' with the bound morpheme 'ex-'. 
Thirdly, the 'pseudomorphological' type of ambiguity was also featured 
in the riddles, whereby an actual word is deliberately confused with part 
of another word (though not a morpheme). This type of ambiguity is 
involved in the riddle, 'What did the hat say to the scarf? (You stay here 
and hang around and I'll go on a head)', Here, the word 'head' is 
deliberately confused with the phonological sequence 'head' in the word' 
ahead'. This second phonological string cannot be described as a 
morpheme, as it is not used as a unit which gives meaning in any other 
word. 

About a fifth of the riddles collected were metaphorically-based and 
these mainly involved the use of conventional polysemy (i.e., metaphors 
already sanctioned in the language). Children from all of the age groups 
made use of this type of ambiguity in their riddles, although its use 
increased amongst the 10-year-olds and then decreased amongst the 11-
year-olds. Examples of riddles which might be described as involving 
conventional polysemy include the following: 

Why did the chicken cross the road? To show he had guts. (Girl, 10 
years). 

Why did the burglar buy a surfboard? Because he wanted to stay on 
the criminal wave. (Boy, 10 years) 

Why did the boy bury his clock? Because the batteries were dead. 
(Girl, 11 years). 

57 



 

STAUNTON 

Only a couple of instances of metaphorically-based riddles 
involving 'radical' or novel polsemy were recorded. One 8-yearold girl 
supplied a riddle which involves comparison (between an egg and a 
tennis ball) This riddle involved partial obscuring of the answer, but 
would be very difficult for the riddlee to guess, based on everyday 
knowledge of the world: 

What s round and yellow,  
Laid in a nest 
But won't hatch like the rest?  
A tennis ball. 

A version of the classic example of the use of metaphor in riddling, 
the Riddle of the Sphinx, was supplied by one 9-year-old boy: 

What has four legs in the morning, two legs at midday and in the 
evening has three legs? 
Answer: 'It’s the three stages of man. He crawls in the morning, then 
walks and then gets a walking stick '. 

Whilst the expectation set up in the riddle image is that some type of 
mythical creature is being described, the metaphor in fact involves a 
comparison with the life stages of a human being. 

A feature which appeared almost exclusively in the riddles of the 
oldest children in the study (the 11-year-olds) was the use of anomaly. 
An early example of a riddle which might be said to involve anomaly, in 
the form of a contradiction based on knowledge of the world, was the 
following 'puzzle', given by an 8-year-old boy: 

Greek Riddle: This Greek king came into his castle and these people 
came along and tried to take him. But before they could take him, he 
took poison. 

Why did they still take him? 

Answer: Greek kings take small bits of poison all the time so that if 
they have to take poison it won't harm them. And that’s why they still 
took him—because he was still alive. 

58 



 

RIDDLE USE AND COMPREHENSION 

More typical examples of the use of anomaly in riddling were given 
by the 11-year-olds: 

What gets bigger the more you take out of it? A hole. 

What gets wetter the more it dries? A towel. Girl, 11 years of age) 

Riddle: What’s as green as grass, as red as a rose, white as snow, 
black as coal? 

Answer: A blackberry. When it’s not ripe it’s green. It flowers white. 
When it's not quite right it’s red and when it’s ripe it turns black. (Boy, 
11 years) 

The 11-year-olds in the present study also provided a few riddles 
which involved the generation of 'unnatural facts'. The answers to these 
riddles usually involved the naming of imaginary creatures whose 
features were a combination of real life attributes and imaginary ones. 
Riddles of this nature included the following: 

What do you call a fly without wings? A walk. (Boy, aged 11 years) 

What has sixteen wheels and pink spots? An elephant with the measles 
on rollerblades. (Girl. aged 12 years) 

The riddling device known as 'catching' was also employed most 
often by the 11-year-olds in this study. A characteristic of catch riddles 
is that the riddle question appears to be a conundrum and the riddlee is 
tricked into thinking that an ingenious answer is required (Opie and 
Opie, 1959). The answer, however, does not involve any type of play on 
words and is instead a plain fact, stated in a very ordinary manner. It is 
for this reason that these riddles can be described as 'victimising' the 
riddlee (McDowell, 1979), as illustrated in these examples from the cur-
rent study: 

What’s pink and fluffy? Pink fluff 

Why did the tortoise cross the road? He wanted to get to the Shell 
station. (Boy, aged 10 years) 

59 



 

STAUNTON 

Why did the turkey cross the road? It was the chicken’s day off. (Girl, 
aged 11 years) 

Summary: Children's Use of Riddles 

The riddles in this sample involved three main types of construction. 
Over half the riddles collected made use of a structure based on 
attribution. Almost a third of the riddles were concerned with causality 
and took the form of a 'Why?/ Because' question and answer sequence. 
The remaining riddles (just over a quarter of the sample) involved the use 
of the structure, 'What did the X say to the YT, whereby inanimate 
objects are featured in imaginary conversational situations, and the two 
meanings of the a well known figure of speech are exploited. A 
quantitative summary of the types of riddle construction used by children 
from each of the age groups is shown in Table 8. 

A number of types of riddling device were used in this sample to 
provide ambiguity or confusion. A few examples of the 'descriptive 
routine' which does not contain a block element were given by an 8-year-
old girl in the study. Simple lexical ambiguity was the most commonly 
used source of ambiguity in the sample, followed by the use of metaphor 
known as conventional polysemy. 

The younger children were more likely to use novel metaphors than 
the older group. The use of metaphoric terms already in existence in the 
language (conventional polysemy) was favoured by the 8- and 10-year-
olds in the study. The 10-year-olds made more use of metariddles than 
any other age group. The use of phonological ambiguity and 
conventional polysemy decreased at 11 years of age, to be replaced by 
the use of riddles involving anomaly and the use of 'catch riddles'. A 
quantitative summary of the types of riddling device favoured by the 
children in different age groups is given in Table 9 below. 

60 



 

RIDDLE USE AND COMPREHENSION 

TABLE 8: The types of riddle structure used by children from each 
age group (n = 24). 

 8 years 9 years 10 Years 11 Years 
Attribution 9 3 5 11 
Causality 4 5 3 1 

'What did the 
X say to the Y?' 3 2 2 1 

TABLE 9: The numbers of riddles involving each type of rid- 
dling device used by children from different age groups (n =24). 

 8 years 9 years 10 years 11 years 
Descriptive 
Routine 4 0 0 0 

Phonological 
Ambiguity 3 4 5 1 

Morphological 
Ambiguity 0 2 2 0 

Radical Polysemy  1 0 0 

Conventional 
Polysemy 3 1 4 1 

Catch Riddles 1 0 1 4 

Metariddles 1 1 4 2 

Anomaly / Use of 
unnatural facts 1 0 0 6 

3.5 A Comparison of the Performances of Boys and Girls in the 
Study 

The performance of the boys and girls on each of the three tasks used in 
the study was compared in both quantitative and qualitative terms. 
Whilst the overall performance of the two groups was found to be 
similar at every age level, there were some areas where slight differences 
were noted. 

61 



 

STAUNTON 

(1) Answering Metaphoric Riddles 

The types of answer given by the girls and boys to the 'true' riddles (1-6) 
were firstly compared using a quantitative type of analysis. As described 
earlier, the three types of answer which were given on this task were the 
'target' answer (the answer which the tester had in mind), an 'alternative' 
answer or a 'Don't know' type of response. When the overall accuracy of 
the two groups of children in answering the riddles was compared, it was 
observed that the girls were slightly more accurate than the boys in 
providing the target answer. The girls gave the target answer for 22% of 
the riddles, compared with 17% for the boys. The boys were more likely 
overall to give an 'alternative' type of answer than the girls. 

The types of answer given by the two groups of children within each 
age group were also examined. Whilst in the 8 and 11 year age groups, 
the girls were found to have given more target answers than the boys, the 
9- and 10-year-old girls and boys were found to have performed equally 
well. The 8-, 9- and 11-year-old boys gave approximately 15% more' 
alternative' answers than the girls in these age groups, though this pattern 
was reversed amongst the 10-year-olds. 

Secondly, a qualitative type of analysis was used to compare the 
validity of the' alternative' types of answer given by the boys and girls. 
The answers were examined in terms of their ability to complete the 
condition of a 'partially obscured semantic fit' between the riddle image 
and the answer. As described earlier, six different types of answer were 
identified from the overall sample of alternative answers. Answers which 
were labelled 'anomalies' and 'literal answers' were given with equal 
frequency by the boys and the girls. These were answers which were not 
considered to complete the riddles, as they usually belonged to the same 
superordinate semantic category as the item implied in the riddle image. 
A similar group of answers was those which involved a 'change of state' 
so that the human or animal in the answer was made to fit the description 
in the riddle image. These were found to be used more often by the boys 
than the girls. The girls, however, were found to have given more 
'replica-type' answers overall than the boys. 

62 



 

RIDDLE USE AND COMPREHENSION 

Of the answers which were judged as completing the riddles (i.e. 
those involving metaphor), more answers based on a novel type of 
metaphorical similarity were used by the girls than the boys. However, 
the boys were found to have given a third more answers which were 
judged as completing the riddles than the girls. They also gave twice as 
many answers which were based on conventional polysemy as the girls. 

(2) Comprehension of the Metaphoric Riddles 

The performance of the boys and girls on the task which probed for 
comprehension of the riddles was also compared. Whilst the 
performance of both groups differed slightly at each age level, no 
obvious developmental trends were apparent. The accuracy of the boys 
in the 8 and 9 year age groups in recalling the grounds of each riddle, for 
example, was slightly better than that of the girls. The opposite was the 
case, however, amongst the 10- and 11-yearolds. 

Only at the age of 10 years was there a large discrepancy in the 
performance of the two groups, where the girls comprehended the 
grounds of 81 % of the riddles compared with the level of accuracy of 
69% obtained by the boys. The latter score, however, is not in keeping 
with the overall trend of improved comprehension with increasing age 
which was displayed by the other age groups. The proportion of 
'alternative responses' given by boys and girls on the probe task was also 
found to be similar overall. The only exception to this finding was 
amongst the 8-year-olds, where the girls were far more likely than the 
boys to compare the items in the probe question on grounds other than 
those in the original metaphor. 

The children's comprehension of the riddles was also examined in 
terms of the effect of the type of metaphoric similarity (i.e. sensory or 
nonsensory) on comprehension. With the exception of one age group, 
both the girls and the boys were found to have greater comprehension of 
the riddles grounded in sensory similarity than the nonsensory riddles. 
Only the 10-year-old boys showed greater comprehension of the 
nonsensory than the sensory metaphors, though their level of 
comprehension of both types of riddle was again not in keeping with the 
overall pattern shown by the other groups. 

63 



 

STAUNTON 

A qualitative comparison of the girls' and boys' explanations of the 
grounds of the nonsensory metaphors was also used. The pattern of 
increasing depth of comprehension of the nonsensory metaphors with age 
was found to be similar for both groups of children. The ability to 
verablise both meanings of the relational types of metaphor was found to 
emerge at 8 or 9 years amongst the boys and the girls. Similarly, the 
ability to explain the grounds of the psychological-physical types of 
metaphor emerged slightly later overall in both groups. 

A qualitative examination of the' alternative responses' supplied by 
the children showed similar patterns overall in the types of creative 
comparisons used by the boys and the girls at each age. Whilst the 
comparisons based on sensory grounds were characteristic of the younger 
children, the older groups tended to use relational comparisons, as 
described earlier. Neither group used psychological-physical types of 
comparison in their creative responses. These patterns were generally 
consistent for both boys and girls. 

(3) Use of Riddles 

Due to the small size of the sample of riddles collected from the children, 
the riddles were grouped in four ways for the purposes of comparison of 
the patterns of use of the boys with those of the girls. The riddles were 
categorised according to the type of riddling device involved in each 
case. The four overall categories of riddle were those based on 
grammatical ambiguity, riddles involving metaphor, catch riddles and 
metariddles, and lastly, riddles involving anomaly or the use of unnatural 
facts. 

The first finding was that overall the boys contributed a greater 
number of 'riddles proper' than the girls. Whilst twenty-six of the riddles 
in the sample were collected from the boys, the girls contributed twenty-
one riddles. However, an additional four 'descriptive routines' were 
contributed by one of the 8-year-old girls in the sample. Secondly, the 
boys were found to have used twice as many riddles involving 
grammatical ambiguity as the girls. Whilst the girls were most likely to 
use this type of riddle at the age of 8 years, its use was favoured by the 
boys at the ages of 9 and 10 years. 

64 



 

RIDDLE USE AND COMPREHENSION 

There was little difference, however, between the boys and girls in 
their use of the other riddling devices. Whilst the boys were found to 
have used marginally more metaphoric riddles than the girls in the study, 
the girls made slightly more use of catch riddles, metariddles and riddles 
involving anomaly. Both boys and girls used more riddles involving 
conventional polysemy as a block element than radical polysemy. The 
use of catch riddles, metariddles and riddles involving anomaly was also 
favoured by the older children in both groups. 

Summary: A Comparison of the Performance of Boys and Girls in 
the Study 

The performance of the girls and boys on each of the three tasks in the 
study was compared in both quantitative and qualitative terms. 

(1) Summary: Answering Metaphoric Riddles 

The types of answer given by the girls and boys to the 'true' riddles (1-6) 
were firstly compared from a quantitative perspective. Whilst the girls 
were found to have given slightly more target answers overall, the boys 
were more likely to give an 'alternative' type of answer. Secondly, the 
answers were compared in terms of their ability to satisfy the condition 
of a partially obscured semantic fit with the riddle image. Overall, the 
boys gave more answers judged as completing the riddles than the girls, 
and they also used twice as many answers based on conventional 
polysemy. The girls were more likely to base their answers on .a novel 
type of metaphorical similarity. The types of riddle answer given by the 
girls and boys are compared in Tables 10 and 11 below: 

TABLE 10: An overall comparison of the types of riddle answer 
given by boys and girls in the study (n=24). 

 Target Answers Alternative Answers 'Don't Know' 
Responses 

Girls 22% 33% 45% 
Boys 17% 43% 40% 

65 



 

STAUNTON 

TABLE 11: A comparison of the types of answers given by the girls and 
boys in the study to the riddles based on sensory similarity (n = 24). 

8 Years 
girls boys 

9 Years 
girl boys 

10 Years 
girls boys 

11 Years 
girls boys 

Target        
Answers  17% 5% 0% 0% 28% 28% 44% 33% 

Alternative        

Answers  39% 55% 44% 61% 28% 17% 22% 39% 
'Don't Know' 
Responses        

 44% 39% 66% 39% 44% 55% 34% 28% 

(2) Summary: Comprehension of Riddles (Boys and Girls) 

The performance of the boys and the girls on the task which probed for 
comprehension of the metaphoric riddles was firstly compared in 
quantitative terms. Whilst the level of accuracy obtained by both groups 
differed slightly at each age level, no obvious developmental trends were 
apparent. The only age at which there was a large difference in the 
performance of the two groups was at 10 years, where the level of 
accuracy obtained by the boys was much lower than that of the girls. 
With the exception of the 10-year-old group just mentioned, both the 
girls and boys were found to have greater comprehension of the riddles 
grounded in sensory, rather than non sensory similarity at each age level. 

A qualitatative comparison of the children's explanations of the 
nonsensory metaphors found that the pattern of increasing depth of 
comprehension with age was similar for both groups. The ability to 
verbalise the grounds of the relational and psychological-physical 
metaphors was in general found to emerge at the same ages for both boys 
and girls. The proportion of 'alternative responses' given by boys and 
girls on the probe task was also found to be similar overall. Comparisons 
based on shared sensory features were in general typical of the younger 
children in each group, whilst relational metaphors were used from the 
age of 9 years. A quantitative comparison of the types of response given 
on the probe task by the girls and boys at each age level is given in Table 
12. 

66 



 

RIDDLE USE AND COMPREHENSION 

The influence of the type of metaphoric similarity on the level of 
accuracy of the two groups is summarised in Table 13. 

TABLE 12: The types of response given by boys and girls in 
different age groups on the task testing for comprehension of the 
metaphoric riddles (n = 24). 

8 Years 
girls boys 

9 Years 
girl boys 

10 Years 
girls boys 

11 Years 
girls boys 

Target 
Answer
s  

69.5 75% 78% 83.5% 81% 69.5% 86% 80.5% 

Alterna
tive 
Answer
s 

19.2% 0% 19.5% 14% 11% 14% 8.5% 8.5% 

'Don't 
Know' 
Respon
ses 

11. 3% 25% 2.5% 2.5% 8% 16.5% 5.5% 11% 

TABLE 13: Accuracy of boys and girls in comprehending riddles 
involving sensory and nonsensory metaphor (as judged by the use of the 
probe question task) (n = 24). 

8 years 9 years 10 years 11 years 
girls boys girls boys girls boy girls boys 

Sensory 
Similarity 89% 83%       83%   89%        84%  67% 94% 83% 

Nonsensory 
Similarity 83% 89%       83%   89%        84%  67% 94% 83% 

(3) Summary: Use of Riddles (Boys and Girls) 

The boys were found to have contributed a greater number of 'riddles 
proper' overall than the girls. They were also found to have used twice as 
many grammatically-based riddles as the girls in the study. There was 
little difference between the boys and girls in their use of the other 
riddling devices. Both boys and girls used more riddles involving 
conventional polysemy as a block element than radical polysemy. The 
use of catch riddles, metariddles and 

67 



 

STAUNTON 

riddles involving anomaly was also favoured by the older children in 
both groups. A quantitative sununary of the types of riddling device used 
by the boys and girls in each age group is given in Table 14 below. 

TABLE 14: The numbers of riddles involving each type of riddling 
device used by girls and boys from each age group (n = 24). 

8 years  
girls boys 

9 years  
girls boys 

10 years 
girls boys 

11 years 
girls boys 

Grammatically-
Based 

2          1 1        5 1        5 1       0 

Metaphorically-
Based 

3          1          0 2 2 2 0        1 

Catches/ 
Metariddles 

1          1 0         1 3         1 3 3 

Anomaly/ 
Unnatural Facts 

0          1 0 0 0 0 2 4 

DISCUSSION 

4.1 Introduction 

This study sought to explore the development of riddle use and 
comprehension in Irish school-aged children. This involved the 
examination of: 

a) the types of answer which children from different age groups 
gave to the metaphoric riddles 

b) the type of response given by children from different age 
groups to a task probing for comprehension of metaphoric 
riddles 

68 



 

RIDDLE USE AND COMPREHENSION 

c) the effect of the type of metaphoric similarity on com-
prehension of these riddles 

d) the types of riddle which were contributed by children from 
different age groups. 

The following four questions had been posed by the researcher: 

1. Are there developmental patterns in the types of  
answer which children give to metaphoric riddle questions? 

2. Does comprehension of metaphoric riddles improve with 
age? 

3. Do children use certain types of riddles before they use 
others? 

4. Are there differences in the developmental patterns displayed 
by boys and girls on these tasks? 

Each of these questions will be examined in terms of the results 
which were obtained in this study. The overall conclusions which may 
be drawn from these results will then be outlined. Finally, the limitations 
of the study will be set out and suggestions for further research will be 
made. 

4.2 Is There a Developmental Pattern in the Types of Riddle 
Answers Used by Children? 

The approach of children to the answering of true riddles is an area 
which, up to this point, has not been investigated by researchers. 
However, the results of the current study would suggest that the tendency 
of providing canonical answers to such riddles is present in some 
children from the age of 8 years. Furthermore, it appears that this 
tendency continues to become more marked between the ages of 8 and 
11 years. 

The first finding of note in this study was that the tendency to give 
answers which are typical of those found in riddle books increased with 
age. Secondly, the creative answers which were offered by the older 
children in the study were found to be more 

69 



 

STAUNTON 

canonical than those offered by the younger groups. Both findings would 
suggest that children's awareness of the conventional requirements of the 
riddling form gradually increases with age. 

As stated above, it was found in the current study that the tendency 
of children to give a canonical riddle answer seems to increase overall 
with age. However, the 9-year-olds in the study were an exception to this 
overall pattern. Whilst a small proportion of the 8-year-olds provided the 
answer which the tester had in mind, none of the answers given by the 9-
year-olds were of this nature. A possible explanation for this finding is 
the small size of the sample of children used in the study (six in each age 
group). The abilities of the six particular children in this group may not 
be representative of the children in their age group in general. 

The high proportion of 'alternative' answers given by children in 
every age group provides support for the notion put forward by Kaivola-
Bregenhøj (1996) that riddles are solved through a process of 'creative 
guessing'. The qualitative examination of this type of answer suggests 
that children's awareness of the conventional requirement of a partially 
obscured semantic fit between the riddle image and answer seems to 
increase with age. The riddle answers which were judged as fulfilling this 
condition were seen as satisfying three criteria. They were firstly 
members of a semantic category opposite to that of the item implied in 
the image, they secondly shared some minor characteristics with the item 
in the riddle image, and they were thirdly prototypical members of the 
category to which they belonged. 

The results of this study suggest that there may be a developmental 
sequence in which children develop awareness of the conventional need 
to fulfill each of these criteria. The' anomaly' answers which were used 
by the younger children in the study, for example, were found to fulfill 
none of the criteria necessary for a partial obscurity of semantic fit. These 
answers generally belonged to the same semantic category or 'set' as the 
item in the riddle image. They tended not to be prototypical members of 
that category and were found not to fit the description given in the image, 
even in a literal sense (e.g. 'What can howl but has no mouth? A frog '). 
However, in each case the children had created an explanation to make 
the answer fit the description in the image. This suggests an emerging 
awareness of the need for the riddle image to have features in common 
with its answer. 

70 



 

RIDDLE USE AND COMPREHENSION 

The 'literal' answers given in the study were found to fulfill the 
criteria of having features in common with the item implied in the riddle 
image. This type of answer was mainly used by the 8and 9-year-olds in 
the study, with some instances amongst the older children. Such answers 
do not involve the use of metaphor, however, as they belong to the same 
semantic category as the item implied in the image (i.e. animate). About 
half the answers given could not be considered prototypical members of 
this category (e.g. 'tadpoles', 'a bat'). Awareness of the need for the riddle 
answer to have features in common with the riddle image would 
therefore appear to be the first of the three riddle-answering skills to 
emerge. 

A similar group of answers were those which fitted the riddle image 
literally but which involved a 'change of state' (e.g. death, the use of 
magic) so that the animate entity was made to fit the description in the 
image. Examples of this type of answer were found in all age groups. 
The use of such a 'change of state' might indicate an emerging awareness 
that riddle answers are generally obscured from the sight of the riddlee in 
some way and that the use of logic might be necessary in order to solve 
the riddle. However, these answers did not fulfill two of the criteria 
necessary for riddle answers. They belonged to the same category as the 
item implied in the image and were not prototypical members of this 
category due to the special conditions displayed by each of the members 
(e.g. 'a blind man'). 

The group of 'replica-type' answers which were given by the 9- and 
10-year-olds, in the study were found to fulfill two of the three criteria 
necessary for a partial obscurity of semantic fit. These answers belonged 
to the inanimate category and had characteristics in common with the 
item implied in the riddle image. The fact that these answers were 
typically used by a slightly older group of children, suggests that the 
awareness of the need to give an answer which is from an opposite 
semantic category might be the second of the three riddle-answering 
skills to emerge. The 'replica-type' answers were not judged as being 
metaphorical, however, as they shared too many characteristics with the 
item in the riddle image. The answers therefore tended not be 
prototypical members of the inanimate category. Aitchison (1994: 150) 
points out that although the items compared in a good metaphor must 

71 



 

STAUNTON 

have some characteristics in common, 'the items must not be too similar 
because the metaphor will be incomprehensible'. 

The answers which were judged as having the most validity as 
canonical riddle answers were those which involved the use of metaphor 
and also fulfilled the three criteria outlined earlier. Whilst some 
metaphorical answers were given by the younger children in the group, 
these tended to involve the use of 'radical polysemy'. These answers 
therefore involved the use of a novel metaphorical similarity which an 
individual child had noticed. Some of these answers were not considered 
to be prototypical members of the inanimate category (e.g. 'The safety 
sticker on a lamp') and would be very difficult for a riddlee to guess. The 
metaphorical answers which were most typical of the 10- and 11-year-
olds were those involving conventional polysemy. These metaphorical 
answers were considered to be prototypical members of the inanimate 
category, as they involved comparisons which are already sanctioned in 
the language. It would appear from these responses that the awareness of 
the need to give a prototypical riddle answer might be the last of the 
three conventional riddle-answering skills to develop. 

The overall findings of this study would suggest that the awareness 
of the conventional requirement of satisfying the condition of a partially 
obscured semantic fit may emerge in three general developmental stages. 
Whilst there was some individual variation in this pattern amongst the 
children in this study, an overall developmental sequence was observed. 
The tendency of giving an answer which has characteristics in common 
with the item in the riddle image was found to be present in children 
aged 8 and 9 years. Whilst the tendency of giving an answer from the 
opposite semantic category was displayed by a few of the younger chil-
dren, this feature typically emerged at 9 or 10 years of age. The tendency 
of giving an answer which satisfied all three answering conditions, was 
most frequently observed amongst the 10- and 11-year-olds in this study. 
This would suggest that, in general, the tendency of generating creative 
answers which satisfy the requirement of a partial obscurity of semantic 
fit emerges between the ages of 10 and 11 years. 

72 



 

RIDDLE USE AND COMPREHENSION 

4.3 Does Comprehension of Metaphoric Riddles Improve with Age? 

Whilst no other study has specifically examined children's com-
prehension of metaphoric riddles, the development of comprehension of 
grammatically-based riddles and the development of metaphor 
comprehension are both areas which have been investigated by 
researchers. The studies of children's riddle comprehension have 
produced conflicting findings. While Prentice and Fathman (1975) and 
Shultz (1974) reported a steady improvement in riddle comprehension 
with age, Fowles and Glanz (1977) found no such relationship. There is 
much evidence in the literature, however, to suggest that the 
comprehension of metaphors improves with age (e.g. Asch and Nerlove, 
1960, cited in Billow, 1977; Gentner and Stuart, 1983, cited in Winner, 
1988; Pollio and Pickens, 1980; Pollio and Pollio, 1979; Winner, 
Rosenthiel and Gardner, 1976). 

The results of the present study suggest that there is an overall 
improvement in the comprehension of metaphoric riddles between the 
ages of 8 and 11 years. The greatest increase in the level of 
comprehension occurred between the ages of 8 and 9 years. The 
performance of the 10-year-olds in this study, however, was found.to be 
poorer than that of the 9-year-olds. Fowles and Glanz (1974) reported a 
similar irregularity in the developmental patterns of riddle 
comprehension displayed by the children in their study. They concluded 
from their results that riddle comprehension was not related to age, but 
to reading ability. The children in their study who displayed superior 
riddle comprehension were found to be those who were also rated as 
successful readers by their teachers. They state that the ability to attend 
to the surface properties of the language, rather than focusing solely on 
the literal interpretation of the words, is one which is necessary for 
riddle comprehension and is also essential to reading. Unlike the riddles 
used in this study, the riddles used by Fowles and Glanz (1977) were 
mainly grammatically-based. The ability to extract nonliteral meaning is 
none the less necessary for comprehension of both types of riddle. Had 
an assesssment of reading ability been used with the children in the 
current study, it might have been possible to establish whether such a 
link between metaphoric riddle comprehension and reading ability, 
rather than age, was in existence. 

73 



 

STAUNTON 

The overall pattern of improved comprehension with increasing age, 
which was displayed by the other groups in the current study, would 
suggest, however, that the performance of the group of 10-year-olds in 
this study may not have been representative of that children from this 
age group in general. Had a larger sample of children been included in 
the study, a more even developmental pattern of improving 
comprehension might have emerged. Such a pattern of development was 
reported in the studies by Prentice and Fathman (1975) and Shultz 
(1974). 

The Language Experience Hypothesis 

The finding of this study that even 8-year-old children displayed 
comprehension of metaphoric language contradicts the cognitive pre-
requisite hypothesis put forward by Piaget (1926, cited in Kamhi and 
Friemoth Lee, 1988). According to this hypothesis, the attainment of the 
stage of formal operational thought is a prerequisite for comprehension 
of figurative forms of language. 

Whilst no assessment of cognitive abilities was administered to the 
children in this study, the formal operational stage typically does not 
begin until the age of twelve years. It might therefore be fair to assume 
that the younger children in this study would not yet have attained this 
cognitive stage of development. This overall finding that metaphor 
comprehension emerges prior to the formal operational stage of thinking 
concurs with those of Billow (1975), Cometa and Eson (1978) and Smith 
(1976). 

The results of this study provide some support, however, for the 
language experience hypothesis proposed by Ortony, Turner and Larson-
Shapiro (1985). According to this theory, children's comprehension of 
figurative language develops gradually through meaningful exposure to 
such language. As they grow older, children have more exposure to 
metaphoric language and more opportunity to test out their 
understanding. The depth of comprehension of metaphors would 
therefore be expected to increase with age. Such a pattern of 
development was noticed amongst the children in this study between the 
ages of 8 and 11 years. A number of researchers have noted this pattern 
of an increasing level of metaphor comprehension as children grow older 
(Asch and Nerlove, 1960, cited in Billow, 1977; Gentner and Stuart, 
1983, 

74 



 

RIDDLE USE AND COMPREHENSION 

cited in Winner, 1988; Pollio and Pickens, 1980; Pollio and Pollio, 
1979; Winner, Rosenthiel and Gardner, 1976). 

4.3.1 The Effect of the Type of Similarity 

There is some controversy in the literature regarding the effect of the 
type of metaphoric similarity on comprehension. A number of 
researchers have found that metaphors grounded in non sensory 
similarity are more difficult for children to comprehend than those based 
solely on sensory characteristics (e.g. Gentner and Stuart, 1983, 
Mendelsohn, Gardner and Winner, 1981 and Shantiris, 1983, all cited in 
Winner, 1988). A small number of studies have reported no difference in 
the ease of understanding of such metaphors (Kogan, Connor, Gross and 
Fava, 1980, cited in Winner, 1988; Nippold, Leonard and Kail, 1984). 
No studies to date have formally compared children's level of 
comprehension of riddles involving the two main types of metaphoric 
similarity. However, Fowles and Glanz (1974) found that riddles 
involving metaphorical language were more difficult for children to 
comprehend than those involving ambiguous, concrete terms. 

The results of the current study suggest that the type of similarity on 
which a metaphor is based can have an effect on children's level of 
comprehension. The children in this study were found overall to have a 
greater level of understanding of riddles based on a sensory type of 
metaphoric similarity than those involving a nonsensory metaphoric 
comparison. The quantitative results of the study suggest that the 
difference in the level of the comprehension of the two types of metaphor 
was greatest at the age of 8 years. The overall level of understanding of 
the sensory metaphors was found to increase only slightly between 8 and 
11 years. However, comprehension of the nonsensory metaphors was 
found to increase greatly overall between these ages, with the greatest in-
crease in the level of comprehension occurring between the ages of 8 and 
9 years. The level of comprehension of the sensory metaphors which was 
displayed by the 10-year-olds in the group is not in keeping with the 
overall developmental trend in this study. As suggested earlier, the small 
size of the overall sample of children included in the study might have 
contributed to this irregularity in the overall developmental pattern. 

75 



 

STAUNTON 

The findings of this study concur with those of the studies cited by 
Winner (1988) that the comprehension of nonsensory metaphors poses 
more difficulty for children of all ages than the comprehension of 
sensory metaphors (Gentner and Stuart, 1983; Mendelsohn, Gardner and 
Winner, 1981; Shantiris 1983, all cited in Winner, 1988). The qualitative 
analysis of the children's performance on the explanation task also 
suggests some developmental differences in the comprehension of the 
two types of nonsensory metaphor used in this study. Whilst some of the 
younger children in the study were able to explain the grounds of the 
relational metaphors, the ability to explain both senses of the 
psychological-physical metaphors did not appear to emerge until the age 
of 10 years. A similar type of trend was reported by Shantiris (1983, 
cited in Winner, 1988), who found that psychological-physical 
metaphors were more difficult for children of all ages to comprehend 
than those grounded in a relational type of similarity. 

The finding that younger children tend to interpret psychological-
physical metaphors in a literal sense has been reported by Asch and 
Nerlove (1960, cited in Billow, 1977) and Winner, Rosenthiel and 
Gardner (I976). The explanations of the grounds of the psychological-
physical metaphors which were given by the younger children in this 
study also involved a 'literal' type of interpretation. Only the explanations 
given by the 10- and 11-yearolds in this study consistently contained 
reference to both the psychological and physical senses of the terms. 
Children aged 8 and 9 years tended to verbalise only the physical sense 
of terms such as 'bright' or 'hard' in their explanations. However, the 
same children had often responded appropriately to the 'probe' task, 
where they simply had to recall the grounds of the riddles presented ear-
lier. This task frequently required only a single-word answer. This 
overall finding is consistent with that of Asch and Nerlove (1960, cited in 
Billow, 1977) who reported that children between the ages of 7 and 10 
years are aware that dual-fuction terms can be applied in a psychological 
sense to people. The ability to explain both senses of the dual-function 
terms was also found to be consistent only after the age of 10 years in 
their study. Winner, Rosenthiel and Gardner (1976) however, report that 
ahnost a third of the 8year-olds in their study were able to explain the 
psychological meaning such terms. 

76 



 

RIDDLE USE AND COMPREHENSION 

Measurement of Comprehension 

The difference in the performance of the younger children on the 'probe' 
and 'explanation' tasks in the current study provides support for the view 
that explanation tasks can under-estimate children's level of metaphor 
comprehension. A number of the 8and 9-year-olds were able to recall the 
appropriate grounds of the nonsensory metaphors when a single-word 
reponse was required, yet their explanations of the same metaphors 
frequently indicated only a 'literal' type of understanding. Theorists such 
as Vosniadou (1987) have criticised the use of explanation tasks due to 
the significant metalinguitic demands involved in verbalising the 
grounds of a metaphor. The results of the studies which have compared 
children's performance on multiple-choice and explanation tasks suggest 
that the ability to comprehend metaphors develops prior to the ability to 
fully explain them (Gardner, 1974; Winner, Rosenthieal and Gardner, 
1976; Pollio and Pickens). It is therefore likely that the children in this 
study had a greater level understanding of metaphors at every age than 
their performance on the explanation task would suggest. 

Sources of Difficulty with Nonsensory Metaphor Comprehension 

The finding that younger children tend to have a lower level of 
comprehension of nonsensory metaphors was attributed by Keil 
(1979,1985, cited in Winner, 1988) to a lack of knowledge about the 
properties of things in the world. The term 'domain knowledge' is used to 
describe children's knowledge of the properties and functions of the 
items in the world's categories or domains. Keil found that in order for a 
child to comprehend a metaphor, he or she must firstly have knowledge 
of the two domains involved in the metaphor (Keil, 1985, cited in 
Winner, 1988). 

The finding that the younger children in the current study displayed 
less understanding of the nonsensory metaphors than those in the older 
groups might therefore be explained in terms of a limited knowledge of 
particular domains. Younger children may not yet have sufficient 
knowledge of domains such as that of personality traits, for example, and 
therefore tend to misinterpret metaphors involving these domains. The 
language experience 

77 



 

STAUNTON 

hypotheisis (Ortony, Larson and Shapiro, 1985) is also useful in 
explaining this finding. The use of non sensory types of metaphoric 
similarity is typically a feature of the spoken and written language of 
adults and older children (Winner, 1988). Younger children are thus less 
likely than their older counterparts to have meaningful exposure to such 
metaphors on a daily basis. The finding that younger children have 
difficulty with nonsensory metaphors might therefore be attributed in 
part to a limited knowledge of the domains involved and a lack of 
meaningful exposure to this type of metaphor. 

In interpreting the results of this particular study, however, the 
syntactic context in which the nonsensory metaphors were presented 
must also be considered as a possible source of difficulty. A small 
number of researchers have reported that the syntactic presentation of a 
metaphor can influence children's level of comprehension (Nippold, 
Leonard and Kail, 1984; Renolds and Ortony, 1980; Winner, Engel and 
Gardner, 1980). The sensory metaphors in this study were presented in 
the context of a true metaphoric riddle, which generally took the form, 
'What has X but no Y? '.  However, the riddles grounded in nonsensory 
similarity took a greater variety of syntactic forms, such as 'What did the 
X say to the Y? ' or riddles involving 'Why / Because' structures. The pos-
sibility that the syntactic presentation of the nonsensory metaphors might 
have contributed the lower level of comprehension of these metaphors 
must therefore be considered also. 

The studies to date which have reported children's greater difficulty 
with nonsensory, than sensory metaphors have generally involved the 
presentation of such metaphors either in a predicative sentence or in 
isolation. The overall results of this study suggest that this finding also 
holds when metaphors are presented in the context of riddles. A small 
number of riddles involving each of the types of similarity was used in 
the study, however, and the number of children included in each age 
group was also relatively small. Further research on a much wider scale 
is therefore necessary before any general conclusions can be drawn. 

78 



 

RIDDLE USE AND COMPREHENSION 

4.3.2 Remarking on Types of Similarity not Implied in the Original 
Riddles 

A small number of children from each age group in the present study 
were found to give an 'alternative' type of explanation when presented 
with the task probing for comprehension of the metaphoric riddles. The 
use of an alternative type of explanation would suggest that these 
children did not comprehend the grounds of the metaphors which they 
had heard earlier. However, given the great creativity of some of the 
children's responses, the use of an alternative type of explanation might 
simply indicate a greater degree of imaginativity on the part of these 
children. The use of an alternative explanation might therefore be 
attributed in some cases to a different approach to the answering of the 
probe questions, rather than a lack of comprehension of the metaphoric 
grounds. The notion that the children used different approaches in 
completing the probe task is evidenced in the finding that certain 
children were much more likely than others use a creative type of 
comparison. 

The fact that the probe questions were based on the metaphoric 
riddles was not made explicit to the children in the study. Some children 
may therefore have used an alternative type of explanation because they 
were not aware that the probe questions related back to the metaphoric 
riddles. A number of children who gave the target response to the probe 
task also went on to describe other similarities between the items 
mentioned in the probe question. Again, this suggests that the children 
may have been eager to demonstrate their creativity in comparing the 
two items, rather than simply giving the target answer only. 

Whilst the overall number of creative (i.e. alternative) explanations 
given in this study was relatively small, this type of response was given 
with greatest frequency by the 9- and 10-yearolds. The finding that 
children in these age groups were the most likely to be creative in their 
use of metaphoric comparisons contradicts the notion of there being a 
'conventional' stage in children's use of metaphor between the ages of 8 
and 10 years (Winner, 1988). Gardner, Kircher, Winner and Perkins 
(1975) reported that children between these ages in their study tended to 
reject the use of metaphor, preferring instead to use language in a 
conventional, unimaginative way. A similar finding was reported 

79 



 

STAUNTON 

by Pollio and Pollio (1974) when a story-writing task was used. 
However, when Pollio and Pollio presented the same group of 
children with a structured comparison task (similar in format to the 
'probe' task used in this study), the children's use of metaphors was 
found to increase steadily between the ages of 8 and 11 years. The fact 
that the comparison task in the current study was presented in the 
context of an informal riddling session might therefore have encouraged 
the children to be less inhibited in their use of language. 

An examination of the creative metaphoric comparisons which were 
used by the children in this study suggests that there is a developmental 
sequence in the types of metaphoric similarity used. The creative 
comparisons which were used by the 8-year-olds in the study were based 
solely on a sensory type of grounds. Children between the ages of 9 and 
11 years were found to favour a relational type of grounds in the 
comparisons which they used. A few instances of the use of 
psychological-physical types of similarity were reported amongst the 
older children in the study. The order of acquisition of the creative use 
of metaphor, which is suggested by these results, is similar in nature to 
that which was observed in the children's comprehension of metaphoric 
riddles in this study. 

The overall results of the current study suggest that the period 
between 8 and 9 years of age may be a time of rapid development in 
both the ability to verbalise metaphoric grounds and in the 
comprehension of metaphors. The greatest overall increase in the level 
of metaphor comprehension amongst the children in this study (as 
judged by their performance on the probe task) was found to occur 
between the ages of 8 and 9 years. The transition from making creative 
comparisons involving a sensory type of similarity to making 
comparisons grounded in relational similarity was also found to occur 
between the ages of 8 and 9 years of age. The use of metaphor by all of 
the children in the study (and not just those who used a creative 
response) would need to, be examined, however, before overall 
conclusions could be drawn in relation to the development of this skill. 

80 



 

RIDDLE USE AND COMPREHENSION 

4.4 Do Children Use Certain Types of Riddles Before They Use 
Others? 

McDowell (1979) concurs with the observation by Sanches and 
Kirshenblatt-Gimblett (1976, cited in McDowell, 1979) that children's 
level of acquisition of the language seems to be reflected in their verbal 
play. These researchers remark that the focus of children's verbal play 
progresses 'from sound play to sense play, finally culminating in play 
with sociolinguistic rules' (McDowell, 1979: 192). The results of the 
current study suggest that by 8 years of age, children are able to make 
use of riddles involving most types of linguistic ambiguity and also 
riddles which involve a playful violation of sociolinguistic rules. 

The current study differs from that of McDowell (1979) in that it 
involves an examination of riddle use by children from the age of8 years. 
The study by McDowell, however, focused on riddle use in children 
below this age. He concluded from his results that the development of 
riddling skills is virtually complete by this age. The finding of the current 
study that children aged 8 years displayed some use of all the types of 
riddles used by the older groups, would also support this notion. The 
patterns of use which were displayed by the children in the current study 
might therefore be interpreted as reflecting their preferences in using 
riddles, rather than indicating a particular stage of acquisition. 

The results of the present study suggest some overall trends in riddle 
use between the ages of 8 and 11 years. A significant finding of this 
study was that the children favoured the use of riddles proper over the 
use of descriptive routines. McDowell (1979) noted in his study that 
older children continue to use descriptive routines once they have 
exhausted their repertoire of riddles proper. He also noted that the 
routines of the older children in his study generally involved the 
description of exotic, rather than mundane items. The descriptive 
routines which were supplied by the 8-year-old girl in this study also 
involved the use of such items (e.g. an elephant, a peacock). 

It appears from the results of this study that children between the 
ages of 8 and 11 years favour the use of homophony as a block element 
in their riddles. The use of grammatically-based riddles in this study was 
found to increase overall up to the age of 10 years. 

81 



 

STAUNTON 

This finding is contrary to that of McDowell, who found a decrease in 
the use of homophony after the age of 7 years. However, a dramatic 
decline in the use of grammatical ambiguity was noted amongst the 11-
year-olds in the current study. Whilst phonological and morphological 
types of ambiguity were featured in the riddles in this study, syntactic 
riddle strategies were not used. 

McDowell (1979) found that metaphoric riddles were favoured by 
the 7- and 8-year-olds in his study. Both radical and conventional types 
of polysemy were featured in the current study. The use of radical 
polysemy was a feature of the 8- and 9-year-olds but was not used 
beyond this stage. There were therefore few examples in the study of 
children's creative use of metaphor. The use of metaphors which are 
already sanctioned in the language (conventional polysemy) was 
featured in a much larger proportion of the riddles. No overall 
developmental trend emerged in the use of this type of metaphor, 
although it was used by children in all age groups. 

The analysis of the riddles collected in this study suggests that the 
period between the ages of 10 and 11 years is a time of great change in 
the types of riddling device used. Whilst the riddles collected from the 
children in the other age groups largely involved the use of linguistic 
ambiguity (including metaphor), the 11-yearolds in the study preferred 
the use of catch riddles, metariddles and riddles involving anomaly or 
unnatural facts. The finding that children in this age group favoured the 
use of 'catch riddles' is in keeping with the sequence of development 
described by Sanches and Kirshenblatt-Grimblett (1976, cited in 
McDowell, 1979). Catch riddles involve the victimisation of the riddlee 
and can thus be described as violating the sociolinguistic rules of the 
language. The popularity of metariddles at this age might also reflect an 
increased eagerness to 'trick' the riddlee. These riddles frequently 
involve the use of a well-known riddle question to which the riddlee 
might give the traditional answer. The child, on the other hand, is in 
possession of the updated answer which the riddlee has far less chance of 
guessing. 

The finding that children in the 11 year age group strongly favoured 
riddles involving anomaly is an interesting one, as McDowell (1979) 
found that the use of such riddles emerges as early as six years of age. 
Whilst it is difficult to explain the popu- 

82 



 

RIDDLE USE AND COMPREHENSION 

larity of anomaly riddles and riddles involving unnatural facts at this 
age, their use might indicate a general move away from linguistic 
ambiguity as a source of humour. Such a pattern of development has 
been reported by researchers such as McGhee (1979, cited in Nippold, 
1988c) who found that adolescents favour jokes where a greater 
cognitive challenge is involved to jokes involving linguistic ambiguity. 

Types of Riddle Structure 

Whilst five main types of riddle structure featured in McDowell's (1979) 
study, the riddles in the present study made use of three types of 
structure. McDowell found that the structures of attribution, causality 
and instrumentality were the most frequently used in his study. The use 
of the structures of attribution and causality accounted for approximately 
three-quarters of the riddles in the current study. The results of the 
present study suggest that the structure of instrumentality is not a feature 
of the riddling of children over the age of 8 years. 

The remaining riddles in this study involved the use of the structure, 
'What did the X say to the Y? '. However, the structure 'What kind of X is 
a Y? " which featured in McDowell's study, was not used by the children 
in this study. The overall results of the current study suggest that the 
structure of attribution is a consistent feature of the riddling of children 
between the ages of 8 and 11 years. The use of the structure of causality, 
however, appears to decline after the age of 9 years. Whilst the use of 
the riddle from, 'What did the X say to the Y?', was most popular 
amongst the 8-year-olds, the results of this study suggest there is a 
decline in its use after this age. 

4.5 Are There Differences in the Developmental Patterns Displayed 
by Boys and Girls? 

The studies to date which have examined children's use and com-
prehension of both riddles and metaphors have reported no significant 
difference in the performance of boys and girls in these areas. The 
results of this study would concur with this overall finding. However 
slight differences in the performance of the boys and girls were detected 
on some of the tasks in this study. 

83 



 

STAUNTON 

Whilst the performance of the boys and girls on the riddle com-
prehension task was very similar at every age, some differences were 
noted in their performance on the riddle-answering task and in their use 
of riddles. Whilst the girls in the current study were found to be more 
likely to give the target answer on the riddle-answering task, the boys 
were found to have contributed a greater number of 'riddles proper' when 
asked to tell the researcher any riddles which they knew. The alternative 
answers which the boys gave on the riddle-answering task were more 
often found to be canonical than those given by the girls. 

The boys were found to have generated twice as many alternative 
riddle answers based on conventional polysemy as the girls, and were 
also twice as likely to contribute riddles involving grammatical 
ambiguity. The girls were more likely to base their answers on a novel 
type of metaphoric similarity. These last findings might suggest a 
preference amongst the boys for the use of types of ambiguity which are 
already sanctioned in the language, as opposed to the more novel 
approach which appears to have been taken by the girls. It was noted by 
the researcher during the collection of the data that the boys generally 
showed a greater interest the reading of riddle books than the girls in the 
study. The use of grammatical ambiguity is a common feature of the 
riddles in such books. The types of response which were given by the 
boys in the study might therefore be related to their greater familiarity 
with this type of ambiguity. A comparison of the performance boys and 
girls would need to be carried out on a much larger scale, however, 
before such findings could be generalised to the population as a whole. 

4.6 Overall Conclusions 

This study set out to examine the development of riddle comprehension 
and use in Irish school-aged children. It was firstly concluded from the 
results of this study that the tendency of children to give canonical riddle 
answers increases between the ages of 8 and 11 years. It also appears 
that there are three general stages in the emergence of children's 
awareness of the conventional requirement of a partial obscurity of 
semantic fit between the riddle question and answer. 

84 



 

RIDDLE USE AND COMPREHENSION 

The results of the current study concur with the widely-reported 
finding that the comprehension of metaphors improves with increasing 
age. The comprehension of metaphors in the context of riddles was found 
to increase overall between the ages of 8 and 11 years. The findings of 
this study provide support for the hypothesis that the comprehension of 
nonliteral language develops gradually through meaningful exposure to 
such language. It also appears from the results of this study that the type 
of metaphoric grounds affects the degree of comprehension of 
metaphoric riddles. The creative use of metaphor which was displayed by 
some children in response to the 'probe' task contradicts the notion of 
there being a 'conventional' stage in metaphor use between the ages of 8 
and 10 years. 

The analysis of the types of riddles collected from the children in 
this study suggests that there is a change at about 11 years of age in the 
types of riddling device which children use. Children up to this age 
appear to favour the use of linguistic ambiguity as a riddling device. 
After the age of 11 years, however, children show a preference for the 
use of catch riddles, metariddles and riddles involving anomaly. 

The performance of boys and girls was compared on all of the tasks 
used in this study. Whilst their performance on the comprehension tasks 
was almost identical, slight differences were observed in the types of 
riddling device which were favoured by each group. Due to the small 
size of this study, however, no overall conclusions about the influence of 
gender on riddling skills can be drawn at this stage. 

4.7 Limitations of this Study and Implications for Further Research 

The main limitation of this study was the small size of the group of 
children included in the study. It is possible that individual differences 
amongst the children may have influenced the results. The children in 
this study were all Irish and all came from a middleclass background. It 
is therefore possible that cultural factors might also influence the results. 
A relatively small number of riddles involving each type of metaphoric 
similarity was also presented to the children. 

85 



 

STAUNTON 

The tasks used in the study would therefore need to be replicated on 
a much wider scale, with children from a variety of cultural 
backgrounds, before any findings could be generalised to the overall 
population. Were the study to be repeated, it might also be of value to 
test out the hypothesis put forward by Fowles and Glanz (1974) that 
riddle comprehension is related to reading ability, rather than age. A 
formal assessment of reading ability would therefore need to be 
administered to the children in the study in addition to the riddle tasks. 

BIBLIOGRAPHY 

Aitchison, J. Words in the Mind. 2nd ed.Oxford: Blackwell, 1994.  
Bee, H. The Developing Child. 6th ed. New York: Harper Collins, 1992  
Billow, Richard, M. "A Cognitive Developmental Study of Metaphor Com- 
 prehension." Developmental Psychology 11[4] (1975): 415-23. 
______. "Metaphor: A Review of the Psychological Literature." Psycho- 
 logical Bulletin 84(1] (1977): 81-92. 
______. "Observing Spontaneous Metaphor in Children." Journal of Ex- 
 perimental Child Psychology 31(1981): 430-45. 
Clark, E. "What's in a Word? On the Child's Acquisition of Semantics in 
 His First Language." Cognitive Development and the Acquisition 
 of Language. Ed. T. E. Moore. New York: Academic P, 1973. 
Cohen, S. "Connecting through Riddles or the Riddle of Connecting." 
 Untying the Knot. Ed. G. Hans-Rokern and D. Shulman. Oxford: 
 Oxford UP, 1996.294-315. 
Cometa, M.S., and M.E. Eson. "Logical Operations and Metaphor Inter- 
 pretation: A Piagetian Model." Child Development 49 (1978): 649 
 59. 
Crossley-Holland, K. The Riddle Book. London: Macmillan Children's 
 Books, 1982. 
Eastman, C. M., and Kallen, J.L. "Structure and Meaning in Swahili 
 Riddles." Journal of American Folklore 92 (1979) 418-44. 
Fathman, R.E., and N.M. Prentice. "Joking Riddles: A Developmental In- 
 dex of Children's Humour." Developmental Psychology 11(1975): 
 210-16. 
Fowles, B., and M. E. Glanz. "Competence and Talent in Verbal Riddle 
 Comprehension." Journal of Child Language 4 (1977): 433-52. 
Franklin, M. B., and S. Barten. Child Language: A Reader. Oxford: Ox- 
 ford UP, 1988. 

86 



 

RIDDLE USE AND COMPREHENSION 

Gardner, H. "Metaphors and Modalities: How Children Project Polar Adjectives 
Onto Diverse Domains." Child Development 45 (1974): 84-91. 

Gardner, H., M. Kircher, D. Perkins, and E. Winner. "Children's Metaphoric 
Productions and Preferences." Journal of Child Language 2 (1975): 125-41. 

Gentner, D. "Children's Performance on a Spatial Analogies Task." Child 
 Development 48 (1977): 1034-39. 
Green, T. A., and W. J. Pepicello. "The Riddle Process." Journal of Ameri- 
 can Folklore 97 (1984): 192-203. 
Hackiyan, A. A Linguistic and Literary Analysis of Old English Riddles. 
 Montreal: Mario Casalini Ltd, 1966. 
Handleman, D. "Traps of Trans-fomation: Theoretical Convergences be- 
 tween Riddle and Ritual." Untying the Knot. Ed. G. Hans-Rokern 
 and D. Shulman. Oxford: Oxford UP, 1996. 38-61. 
Kaivola-Bregenhøj, A "Riddles and Their Us." Untying the Knot. Ed. G. 
 Hans-Rokern and D. Shulman. Oxford: Oxford UP, 1996. 10-36. 
Kallen, J. "Linguistics and Oral Tradition: the Structural Study of the 
 Riddle." Centre for Language and Communication Studies: Occa- 
 sional Paper no.2. Dublin: Trinity College, 1981. 
Kemper, S., and M.L Rice. Child Language and Cognition. Baltimore: 
 University Park, 1984. 
Kittay, E. F. Metaphor: Its Cognitive Force and Its Linguistic Struc- 
 ture. Oxford: Clarendon, 1987. 
Köngäs Maranda, E. "The Logic of Riddles." Structural Analysis of Oral 
 Tradition. Ed. P. Maranda and E. Kongas Maranda. Philadelphia: U 
 of PennsylvaniaP, 1971. 189-232. 
Lakoff, G., and M. Johnson. Metaphors We Live By. London: U of Chi- 
 cago P, 1980. 
Larson-Shapiro, N., A.Ortony, and T.J. Turner. "Cultural and Instructional 

Influences on Figurative Language Comprehension by Inner City Children." 
Research in the Teaching of English. 19 (1985): 2536. 

Lieber, M.D. "Riddles, Cultural Categories and World View." Journal of 
American Folklore. 89 (1976): 255-65.  

McDowell, J. H. Children’s Riddling. London: Indiana UP, 1979.  
Nippold, M.A. "Figurative Language." Later Language Development 
 Ages 9 through 19. Ed. M. A. Nippold. Boston: College Hill, 1988b. 
 179-207. 
______. "Introduction." Later Language Development Ages 9 through 
 19. Ed. M. A. Nippold. Boston: College Hill, 1988a. 1-10. 
_______. "Linguistic Ambiguity." Later Language Development Ages 9 
 through 19. Ed. M. A. Nippold. Boston: College Hill, 1988c. 211 
 24. 

87 



 

STAUNTON 

Nippold, M.A., L. Leonard, and R. Kail. "Syntactic and Conceptual Factors in 
Children's Understanding of Metaphors." Journal of Speech and Hearing 
Research. 27 (1984): 197-205. 

Nippold, M.A., C. L. Taylor, and J.M. Baker. "Idiom Understanding in 
Australian Youth: A Cross-Cultural Comparison." Journal of Speech and 
Hearing Research. 39 (1996): 442-47. 

Nippold, M.A., and M. Rudzinski. "Familiarity and Transparency in Idiom 
Explanation: A Developmental Study of Children and Adults." Journal of 
Speech and Hearing Research. 36 (1993): 728-37. 

Opie, I, and P. Opie. The Lore and Language of Schoolchildren. Oxford: Oxford  
        UP, 1959. 
Ortony, A., and R. E. Reynolds. "Some Issues in the Measurement of 
 Children's Comprehension of Metaphorical Language." Child De- 
 velopment. 51 (1980): 1110-19. 
Ortony, A., R. E. Reynolds, S. Vosniadou, and P. T. Wilson. "Sources of 
 Difficulty in the Young Child's Understanding of Metaphorical Lan- 
 guage." Child Development 55 (1984): 1588-606. 
Pollio, M. R., and J. D. Pickens. "The Developmental Structure of Figu- 
 rative Competence." Cognition and Figurative Language. Ed. R.R. 
 Hoffman and R.P. Honeck. Hillsdale, NJ: Erlbaum,1980. 311-40. 
Pollio, M.R., and H.R. Pollio. "The Development of Figurative Language 
 in School Children." Journal of Psycholinguistic Research 3 (1974): 
 185-201. 
______."A Test of Metaphoric Comprehension and Some Preliminary 
 Data." Journal of Child Language 6[2] (1979): 111-20. 
Rosch, E. "Cognitive Representations of Semantic Categories." Journal 
 of Experimental Psychology: General 104[3] (1975): 192-233. 
Searle, J. "Metaphor." Metaphor and Thought. Ed. A. Ortony. New York: 
 Cambridge UP, 1979. 
Shultz, T.R. "Development of the Appreciation of Verbal Riddles." Child 
 Development 45 (1974): 100-105. 
Smith, J.W.A. "Children's Comprehension of Metaphor: A Piagetian In- 
 terpretation." Language and Speech 19 (1976): 236-43. 
Vosniadou, S. "Children and Metaphors." Child Development 58 (1987): 
 870-85. 
Winner, E. "New Names for Old Things: the Emergence of Metaphoric 
 Language." Journal of Child Language 6 (1979): 469-91. 
______. The Point of Words: Children s Understanding of Metaphor and 
 Irony. London: Harvard UP, 1988. 
Winner, E., M. Engel, and H. Gardner. "Misunderstanding Metaphor: 
 What's the Problem?" Journal of Experimental Child Psychology 
 30 (1980): 22-32. 
Winner, E., A.K. Rosenthiel, and H. Gardner. "The Development of Meta- 
 phoric Understanding." Developmental Psychology 12[4] (1976): 
 289-97. 

88 



 

RIDDLE USE AND COMPREHENSION 

APPENDIX 1 

List of 12 Riddles Used in Task I (Chosen from Children's Riddle 
 and Joke Books) 

Riddles involving a metaphorical comparison based on sensory grounds 

1. What has legs but cannot walk? A table. 
2.What has teeth but no mouth? A comb. 
3.What has eyes but cannot see? A potato. 
4.What goes around all day with its tongue hanging out and sits by your 

bed at night? A shoe. 
5.What runs but never walks? Water. 
6.What can howl but has no mouth? The wind, 

Riddles involving a (nonsensory) metaphorical use of language 

 7.Why did the teacher wear sunglasses to school? Because the class 
 was so bright. 
 8. What happened when the thief fell into the cement? He became a hard- 
 ened criminal. 
 9.Why did the boy throw his clock out the window? He wanted to see 
 time fly. 
10. What do jigsaw puzzles do when they get bad news? They go to pieces 
  / They fallapart. 
11. What did the big telephone say to the little telephone? 'You're too 
 young to be engaged '. 
12. What did the sweeping brush say to the vacuum cleaner? 'Don't you 
 wish people would stop pushing us around?' 

Explanations of the Types of Metaphors Involved in Riddles 1-12 

Riddles involving a metaphorical comparison based on sensory grounds 

Riddles 1-4 are based on a comparison between the physical features of an 
animate entity and those of an inanimate one, 

Riddle 5 is based on a comparison between the movement of a human or animal 
(running) and the movement of a natural object (water). 

Riddle 6 is based on a comparison between the sound of a vocalisation by an 
animate entity and the sound of a natural phenomenon (the wind). 

89 



 

STAUNTON 

Riddles involving a (nonsensory) metaphorical use of language 

Riddles 7 and 8 involve dual-function adjectives which are based on the 
comparison between a physical notion (e.g. brightness or hardness) and a 
psychological one. 

Riddles 9-12 are based on relational types of similarity. 

The comparison involved in Riddle 9 is that between an abstract entity (time 
passing) and the movement of an animate (or inanimate) entity (flight) . 

The comparison on which Riddle 10 is based is that between a physical action of 
falling apart and a psychological state (getting upset). 

The comparison involved in riddle 11 is between the notion of a telephone being 
in use therefore unavailable and someone who is engaged to be married and 
therefore unavailable. 

The comparison involved in riddle 12 is between the physical act of pushing 
something around and the act of bullying someone psychologically. 

Probe Questions Which Followed the Riddles 
(Questions Used in Task 2) 

Riddles involving a metaphorical comparison based on sensory grounds  
(Target answers are in italics) 

1. How (In what way) is a table a bit like a person? It has legs.  
2. How is a comb a bit like a mouth? It has teeth. 
3. How is a potato a bit like a face? It has eyes. 
4. How is a shoe a bit like a mouth? It has a tongue. 
5. How could water be a bit like a person in a race? It can run.  
6. How can the wind be a bit like an owl? It can howl. 

Riddles involving a (nonsensory) metaphorical use of language 

7. How could a person be a bit like the sun? They could be bright.  
8. How could a person be a bit like a rock? They could be hard. 

90 



 

RIDDLE USE AND COMPREHENSION 

Probe Questions Which Followed the Riddles 
(Questions Used in Task 2) 

Riddles involving a metaphorical comparison based on sensory grounds  
(Target answers are in italics) 

1. How (In what way) is a table a bit like a person? It has legs.  
2. How is a comb a bit like a mouth? It has teeth. 
3. How is a potato a bit like a face? It has eyes. 
4. How is a shoe a bit like a mouth? It has a tongue. 
5. How could water be a bit like a person in a race? It can run.  
6. How can the wind be a bit like an wolf? It can howl. 

Riddles involving a (nonsensory) metaphorical use of language 

7. How could a person be a bit like the sun? They could be bright. 
8. How could a person be a bit like a rock? They could be hard. 
9. How can time be a bit like a bird? It can fly. 

10. How could a person a bit like a broken up jigsaw? They've gone to 
 pieces / They're falling apart. 
11. How could a telephone be a bit like a person who's going to get mar- 
 ried soon? It could be engaged. 
12. How could a person be a bit like a sweeping brush? They could be 
 pushed around. 

These questions were presented in random order. 

APPENDIX 2 

The Children's Alternative Answers on the Riddle Answering Task 

1. What has legs but cannot walk? 

8 years: Girls:(None) 
 Boys: Trousers; tadpoles; A person that has a spell put on them. 

9 years: Girls: A doll or a teddy; 
 Boys: A footpath. 

10 years: Girls: A gymnastic horse; 
 Boys: (None) 

11 years: Girls: (None) 
 Boys: A person in a wheelchair. 

91 



 

STAUNTON 

2. What has teeth but no mouth? 

8 years: Girls: A piano; a fork; 
 Boys: Teeth in a box. 

9 years: Girls: False teeth; 
 Boys: A needle (one tooth). 

10 years: Girls: Shark teeth on a piece of string; 
 Boys: A trap. 

11 years: Girls: (None) 
 Boys: A saw (two answers). 

3. What has eyes but cannot see? 

8 years: Girls: A bat; 
 Boys: Joke glasses. 

9 years: Girls: A person in a painting; 
 Boys: A bat; a needle. 

10 years: Girls: A dead fish; Cats' eyes on the road; 
 Boys: A pin (needle). 

11 years: Girls: (None) 
 Boys: A blind man. 

4. What goes around all day with its tongue sticking out and sits by your bed at 
night? 

8 years: Girls: A dog; The safety sticker on a lamp; 
 Boys: A dog. 

9 years: Girls: A dog; 
 Boys: A dog or cat; A rude teddy. A dog. 

10 years: Girls: A dog; 
 Boys: (None) 

11 years: Girls: A dog; Your tie; 
Boys: (None) 

92 



 

RIDDLE USE AND COMPREHENSION 

5. What runs but never walks? 

8 years: Girls: A runner; A cheetah; 
 Boys: Your nose; A car engine; 

9 years: Girls: An ostrich; Your nose; 
 Boys: Runners (shoes); A bicycle or rollerskates. 

10 years: Girls: (None) 
 Boys: A tape recorder; 

11 years: Girls: A car (the engine); 
 Boys: A robot that is programmed to run, not walk. 

6. What can howl but has no mouth? 

8 years: Girls: A radio; 
 Boys: The alarm system (burglar alarm). 

9 years: Girls: A hairdryer; 
 Boys: A frog or toad; A whistle. 

10 years: Girls: (None) 
 Boys: (None) 

11 years: Girls: An Alarm clock; 
 Boys: The radio; The breaks [brakes-ed] on a car;A car. 

APPENDIX 3 

Riddles Collected from Each Class Group 

Third Class (Girls) 
Child No.2 
(1) What did the big biscuit say to the little biscuit? 'Crumbs.' 
(2) What's big and has a very, very big appetite? An elephant. 
(3) What can you get that can be different colours and very soft? A cush- 
 ion. 
(4) What's got two legs, a medium-sized body and a very, very big tail? 
 A peacock. 
(5) What's green and brown and has different colours all over it? A 
 Christmas tree. 

93 



 

STAUNTON 

Child No.3 
(6) What's round and yellow, laid in a nest, but won't hatch like the rest? 
 A tennis ball 
(7) What did the ground say to the rain? 'If you keep this up my name 
 will be mud. ' 
(8) Why does a lion like Christmas? Because of his sandy paws (Santa 

Claus).  
(9) What has keys but can't open any doors? A piano.  
(10) Why did the wolf cross the road? Because the chicken hadn't been 

invented yet. 
Third Class (Bovs) 
Child No.4 
(11) What did the skeleton say to the toilet? 'You can't get anything out 
 of me'. 
(12)Why did the skeleton cross the road? Because he needed to go to The 
 Body Shop. 
(13) What do you call a chimp with a banana in each ear? Anything you 

like—he can't hear you!  
Child NO.5  
(14) Greek Riddle: This Greek king came into his castle and these people 
 came along and tried to take him. But before they could take him, he 
 took poison. 

Why did they still take him? 
Answer: Greek kings take small bits of poison all the time so that if they 
have to take poison it won't harm them. And that's why they 

 still took him-because he was still alive. 
Fourth Class (Girls) 
Child NO.7 
(15) What did the big candle say to the little candle? 'I'm going out to 

night' . 
Fourth Class (Boys) 
Child No. 10 
(16) What did the big chimney say to the little chimney? 'You're to young 

to be smoking'. 
Child No. 11 
(17) Why did the egg go all around the world? Because he wanted to be 

an eggs-plorer! 
Child No. 12 
(18) Why did the egg start crying? Cause it was all cracked up. 
(19) What is a math teacher's favourite tree? A geometree. 
(20) Why did Dennis bring a ladder to assembly? Because his teacher 

wanted him to sing higher. 
(21) What has four legs in the morning, two legs at midday and in the 

94 



 

RIDDLE USE AND COMPREHENSION 

evening has three legs? Answer: It's the three stages of man. He crawls in 
the morning, then walks and then gets a walking stick. 

(22) Why did the tortoise cross the road? He wanted to get to the Shell 
 station. 
(23)What do water and a nose have in common? They can both run if 

they're cold. 
Fifth Class (Girls) 
Child No. 13 
(24)What's black and white and re(a)d all over? A newspaper. Or it could 

be a penguin or a zebra with sunburn. 
(25) What do you call a cucumber with four doors? A slammer. (A slammer 

is a sandwich). 
(26) Why did the chicken cross the road? To show he had guts. 
Child No. 15 
 (27) Riddle: In comes two legs, carrying one leg and puts it on three legs. 

In comes four legs and out goes five legs. Out comes two legs and throws 
three legs at four legs. (And you've got to figure out the answer over 24 
hours). Answer: In comes two legs (a human), carrying one leg (a leg of 
lamb) and she puts it on a three-legged stool. And in comes a four-legged 
dog and takes it outside (and that's the five legs that go out, the dog and the 
leg of lamb). And when two legs sees that the leg of lamb is gone she picks 
up the three-legged stool and goes outside and throws it at the dog. 

Fifth Class (Boys) 
Child No. 17 
(28)Why did the hedgehog cross the road? To show his friends he had 
 guts. 
(29) Why does a golfer wear two pairs of trousers? In case he gets a hole 
 in one. 
(30)Did you hear about the surfing burglar? He stayed on the criminal 
 wave. 
(31) What did the hat say to the scarf? You stay here and hang around and 

I'll go on a head. 
(32)What's a frog's favourite football stadium? Croak Park. 
Child No. 18 
(33) What did the big hand say to the little hand? I’ll be around in an 

hour'. 
(34) Why is the baseball stadium hot after the game? Because all the fans 

have left. 
Sixth Class (Girls) 
Child No. 19 
(35) What has sixteen wheels and pink spots? An elephant with the 

measles on rollerblades. 

95 



 

STAUNTON 

(36)What's black and white red all over? A nun who's fallen down the 
 stairs. 
(37) Supposing, supposing, three men were a-frozing. One died. How many 

were left? None-because you're only supposing. 
Child No. 20 
(38)Why did the turkey cross the road? It was the chicken's day off.  
Child No. 21 
(39) Why did the boy bury his clock? Because the batteries were dead. 
(40) What gets bigger the more you take out of it? A hole. 
(41) What gets wetter the more you dry? A towel. 
(42) What's big, purple and eats spaceships? A big, purple spaceship 

eater. 
Child No. 23 
(43)What do you call a fly without wings.? A walk. 
(44)Riddle: What's as green as grass, as red as a rose, white as snow, 

black as coal? Answer: A blackberry. When it's not ripe it's green. It 
flowers white. When it's not quite right it's red and when it's ripe it turns 
black. 

Child No. 24 
(45) What did the big chimney say to the little chimney? 'You're too young 

to be smoking'. 
(46)What's pink and fluffy? Pink fluff. 
(47)What's brown and sticky? A brown stick. 
(48) What's blue and fluffy? 'Blue fluff?' No, pink fluff with a cold. 

Jokes Collected from the Children 

Third Class (Girls) 
Child No.2 
(1) The Dad asked the girl, 'Do you want any help with your homework?' . 

And the girl said, 'No thanks, I can get it wrong by myself. 
Third Class (Boys) 
Child No.5 
(2) This man was in a pub watching a match. He drank half his glass of 

Harp and put it down on the table. And this little green goblin comes along 
while he's watching the game. And the goblin goes, 'Be, wee, wee, wee, 
wee' (putting his finger in the glass) and fills it up again. Then the man 
comes back, 'Now, don't do that'. And the green goblin comes back and 
goes, 'Be, wee, wee' and fills the glass up again. 

And the man says, 'If you do that again, I'll pull down your yocks and 
and pull your wey off. And the man puts down the drink and watches the 
T.V. again. And the goblin comes back and goes, 'Be, wee, wee', and the 
man goes, 'That's it!'. (Gestures puffing some- 

96 



 

RIDDLE USE AND COMPREHENSION 

thing oft). And the man says, 'Now, how will you go to the toilet? and the 
goblin goes, 'Bee, wee, wee, wee' and the glass overflows. 

(3) If I was a teacher, I'd give my pupils glasses. Get it? My pupils 
glasses. 

Fourth Class (Girls) 
Child No.7 
(4) Knock, knock, 
     Who's there? 
     Boo who? 
     There, there, don't cry. It's only a joke. 
Child No.8 
(5) Knock, knock. 

Who's there? 
Owls. 
Owls who? 
Yes, they do, don't they? 

Fourth Class (Boys) 
Child No. 10 
(6) Knock, knock, Boo. 

Who's there? 
Banana who? 
Knock, knock. 
Who's there? 
Banana. 
Banana who? 
Knock,knock. 
Who's there? 
Orange. 
Orange who? 
Orange you glad I didn't say banana? 

(7) There were three elephants called Shut Up, Mercy and Trouble. And one day 
Trouble went missing. And Shut Up went to the police station and the 
policeman asked Shut Up for his name. So he said 'Shut Up'. And the 
policeman said, 'Are you looking for trouble?'. 'Yes, please'. So the 
policeman said, 'Right then, you're going to get it'. So Shut Up shouted for 
Mercy. And the policeman said, 'It's alright, I wasn't serious'. 

(8) One day a boy wanted to go to the toilet and he asked his teacher if he could 
go. And she said, 'No'. And then she said, 'Say your ABC'. So he said, 
'Abcdefghijklmno_qrstuvwxyz'. 'Where's your 'p' '?. And the boy said, 'It's 
dribbling down my leg'. 

Child No. 12 
(9) This is a joke about Newcastle Football Club, saying that they're 

not very good: 

97 



 

STAUNTON 

'Somebody broke into Newcastle's Trophy Room'. 'Really?'  
'Yeah'. 
'What did they take?' 
'Everything. They're looking for a man with a black and white carpet'. 

Fifth Class (Girls) 
Child No. 13 
(10)There's a little boy and a granny and he's walking down the road when he  

comes across a five pound note. And he says, 'Granny, Granny can I pick the 
five pound note off the ground? And the granny says 'No, because anything 
on the ground is dirty'. And then he comes across a ten pound note and he 
says, 'Granny, Granny can I pick up the ten pound note? And the granny 
says, 'No, because anything on the ground is dirty'. 

And then he comes across a hundred pound note and he says, 'Granny, 
Granny, can I pick up the hundred pound note? And she says, 'No, because 
anything on the ground is dirty'. The granny falls on the ground and she 
says, 'Sonny, sonny, pick me up' and the boy goes, 'No, because anything on 
the ground is dirty'. 

(11) 'Anything I say, you have to say, 'And so did the fat lady'. One day I went 
into a shop... ' 
'And so did the fat lady'. 
'I got bread and milk' . 
'And so did the fat lady'. 
'I paid for the bread and milk'. 
'And so did the fat lady'. 
'I went out of the shop'. 
'And so did the fat lady'. 
'1 bought a balloon'. 
'And so did the fat lady'. 
'The balloon burst'. 
'And so did the fat lady'. 

(12)This man always drove to work on an egg. And one day it wouldn't start. So 
he rang the AA and the AA came out and said, 'Well, we'll take the yoke 
out'. So they took the yoke out and the car started and he drove to work. And 
when he got to work he rang the AA to thank them because his car was all 
white now because they had taken the yolk out. 

Child No. 14 
(13) Mr. McCarthy and Mrs. McCarthy were a family in America. Mrs. 

McCarthy always stayed at home and Mr. McCarthy always went to work. 
And when Mr. McCarthy came home one day he said, 'Did any mail come 
today? And Mrs. McCarthy said, 'No, there's only been female in the house 
all day'. 

98 



 

RIDDLE USE AND COMPREHENSION 

(14) Paddy Englishman, Paddy Irishman and Paddy Scotsman all walked into a 
bar. And one of them said, 'I need to go to the toilet'. So Paddy Scotsman 
walks up to the toilet and he opens the door and he hears this sound: 'I am 
the ghost of the one black eye'. And he runs out really quickly and says, 
'There's a ghost in the toilet!' 

So, Paddy Englishman, thinking he's being really brave, walks in and 
he hears the sound, 'I am the ghost of the one black eye' and he runs out and 
says, 'There really is a ghost in the toilet'. So Paddy Irishman goes in and the 
door opens and he hears the voice, 'I am the ghost of the one black eye'. And 
he says, 'Yeah. And you'll be the ghost of the two black eyes in a second if 
you don't shut up'. 

Fifth Class (Boys) 
Child No. 16 
(15)How do you know a blonde has been at her computer? There's Tippex 
 all over the screen. 
(16) Doctor, doctor, I've got a spanner in my throat.' 'Are you choking?' 
 'No, I'm serious.' 
(17)'Waiter, there's a fly in my bowl'. 'He'll be swimming in the Cup 

tomorrow. '  
Child No. 18  
(18 Paddy English, Paddy Irishman and Paddy Scotsman were all going 

to jail for five years. And they were each allowed to bring five hundred 
things with them. Paddy Englishman brought five hundred bottles of wine, 
Paddy Scotsman brought five hundred cakes and Paddy Scotsman 
[Irishman-ed] brought five hundred cigarettes. When the five years were up, 
Paddy Englishman came out drunk, Paddy Scotsman came out fat and Paddy 
Irishman came out and said, 
'Does anyone have a light? 

Sixth Class Girls) 
Child No. 20 
(19) Paddy Englishman, Paddy Irishman and Paddy Scotsman were out in 

the desert and they each bought one thing in the shop. Paddy Englishman 
bought a bottle of water, Paddy Scotsman bought a baseball cap and Paddy 
Irishman bought a car door. And they each asked each other why they had 
bought these things. Paddy Englishman wanted the water to drink if he got 
hot. Paddy Scotsman wanted the cap to keep the sun out of his eyes. And 
Paddy Irishman wanted the car door so that if it got too hot he could roll 
down the car window. 

Child No. 21 
(19) What goes black and white thump, black and white thump, black and 
 white thump? A nun rolling down the stairs. 
What goes black and white, 'Ha, ha', black and white, 'Ha, ha', black 
 and white, 'Ha, ha'? The nun who pushed her. 

99 



 

STAUNTON 

Sixth Class (Boys) 
Child No. 23 
(20) What do you call a Lada going up a hill? A miracle. 
(21) What do you do if a Kerryman throws a grenade at you? You pull out 
 the pin and throw it back. 
(22) What do you do if a Kerryman throws a pin at you? You run because 
 he probably has the grenade between his teeth. 
(23)How do you drown a Kerryman? Knock on the door of his subma- 

rine.  
(24)Did you hear about the Kerryman's dictionary? It has an index.  
(25)Have you heard about the Kerryman's latest inventions? The water 
 proof teabag and the underwater hairdryer. 
(26) What's black and white and goes, 'Thump, thump, thump, crash?' A nun 

falling down the stairs. 
And what's black and white and goes, 'Ah, ha, ha!' The nun who 

 pushed her. 
Child No. 24 
(27)This idiot is in bed and the phone rings. And the voice says, 'Hello, 

is this 2333964?' And the idiot says, 'No, this is 2333963'.  
'Oh, I'm sorry to get you out of bed'. 'No, it's OK. I had to get out of bed 
anyway to answer the phone. 

100 




