SELIM Carmen Guarddon, Selim 11 (2001-2002): 117—145 THE LOCATIVE USES OF THE PREPOSITION AT IN THE OLD ENGLISH VERSION OF BEDE’S ECCLESIASTICAL HISTORY OF THE ENGLISH PEOPLE. A COGNITIVE APPROACH 1. INTRODUCTION We have chosen the Old English Version of Bede’s Ecclesiastical History as a source to construct our corpus of spatial expressions containing the preposition at because it offers obvious advantages. As a historical work it is remarkably long, which allows us to obtain a reasonable number of examples. Furthermore, its narrative character provides broad contexts against which to draw generalisations, specifically concerning the alternation of different prepositions in expressions which are apparently alike. This factor is particularly relevant when taking into account that preposition alternation usually responds to “unexpected context dependencies” (Herskovits 1986: 15). These context dependencies are connected with the speaker’s – in this case the translator’s – viewpoint of a scene or situation. Since the internal structure of this category affects its distribution with the other basic topological prepositions in and on, the use of the latter will also be considered in our analysis. We have used Miller’s 1890 edition, as it is regarded as the most complete. This edition constitutes almost in its entirety a reproduction of the Tanner Manuscript (T.). This manuscript is kept in the Bodleian Library, Oxford, and was written by five scribes in the first half of the tenth century. Miller has resorted to the following manuscripts to supplement the defects from T.; they are presented in order of preference: C. = Ms Cotton Otho B. XI. British Museum. 0. = Ms Corpus Christi College, Cambridge. Ca. = Ms Kk. Cambridge University Library. Carmen Guarddon 118 The T and C manuscripts are noticeably consistent, thus they are said to resemble the archetype. However, it is still hard to locate this archetype. Miller (1890) explains that although traditionally the translation has been assigned to Alfred there is no trace of a West-Saxon original. The fact is that it is hard to reconcile the Alfredian origin of this version with the Mercian dialect of the text (Cassidy and Ringler 1971: 107; Miller 1890: xxxiii). On the methodological level, we have included all the occurrences of the preposition at with a number of objects that refer to a varied range of spatial categories. For the sake of contextualising the spatial expressions at issue, the whole sentence where the expression occurs has been included in our corpus. The greater attention is paid to the objects of the preposition, as will be seen in the following section, they are principally responsible for segmenting prepositional senses. An analysis of the preposition at in Old English is not as complex as an analysis of the prepositions in and on. In fact, a description of the structure at in this period shows that its usage is quite similar to that of present-day English. In contrast, the prepositions in and on behave quite differently in Old English. This has led some scholars to conclude that their variation responds to local usage (cf. Hols 1971; Miller 1890). The conclusions reached in this analysis serve as a basis for further research into the categorisation of the preposition at in other texts, but we consistently speak of the usage of the preposition at in a particular literary work: the Old English Version of Bede’s Ecclesiastical History of the English People (henceforth: EH). Nevertheless, in order to test to which extent our conclusions can be applied to the use of at in Old English in general, our findings contrast with the analyses of the preposition at in Old and Middle English carried out by Lundskaer-Nielsen (1993) and Lindkvist (1978). Lundskaer-Nielsen attempted to shed some light on the behaviour of the preposition at in Old English and Middle English. In Old English his conclusions derive from an analysis of two extracts of the Anglo-Saxon Chronicle. His initial work covers the years 892-900 and his follow-up covers the Second Continuation of the Peterborough Chronicle 1122-54. Therefore, considering the time-span that separates both sections, his analysis has a diachronic dimension, whereas Lindkvist identifies a great number of occurrences of the preposition at from Old and Middle English texts and compares them with identical uses of its Gothic cognate. Some of The Locative Uses of the Preposition In the OE Version of Bede's EHEP 119 our theoretical claims will be backed up with data based on Lundskaer- Nielsen and Lindkvist’s analyses. 2. SPATIAL COGNITION AND THE SEMANTICS OF PREPOSITIONS The advent of cognitive linguistics has brought about a remarkable interest in the semantic properties of prepositions. Linguists have focused on the ability of these lexical elements to categorise situations over a wide range of domains (space, time, cause, feelings, mental states, etc.). Decades before cognitive linguistics came onto the stage, some grammarians had already verified the spatial origin of prepositions (Brøndal 1950). Cognitive linguists have taken a step further in this direction and argue that prepositional polysemy constitutes solid evidence of the metaphorical operations that underlie human categorisation.1 Specifically, Lakoff and Johnson (1980) showed that spatial metaphors are derived from the bodily basis of experience, and organise our knowledge in other domains. The common categorisation in distinct domains is grounded on the existence of image-schemata (Johnson 1987). Image-schemata are a means of structuring particular experiences, in order to endow our perceptions and conceptualisations with connectedness. The internal structures of image- schemata show a high degree of abstraction; this fact is consistent with the object idealisations involved in the spatial relationships encoded by prepositions. It must be borne in mind that image-schemata are abstract constructs that extrapolate rich patterns of experience. This is demonstrated by Johnson himself when he proposes a representation of the containment schema that underlies the uses of the preposition in, both locative and abstract: Figure 1. The image-schema underlying the uses of the preposition in according to Johnson (1987: 23) 1 There also exist results from neuroscientific research that prove that language was originally aimed at expressing spatial relationships (O’Keefe & Nadel 1978). X Carmen Guarddon 120 This idealisation motivates the use of the same preposition in a number of related but distinct ways, as suggested by Herskovits (1986). Herskovits defines the meaning of prepositions according to an ideal meaning and a number of use-types. The ideal meaning Herskovits suggests for the preposition at is ‘a point that coincides with another’. The word “point” in the ideal meaning of the preposition at indicates, Herskovits explains, that the categorisation of spatial relations responds to geometric conceptualisations of both the subject and the object of the preposition. Therefore, terms that denote an entity that cannot be viewed as a point cannot be the subject of this preposition: * “the road is at the park.” The subject and object of the prepositions are syntactic notions that mirror the order that the relevant entities occupy with respect to these elements; the subject precedes whereas the object follows the preposition.2 Nevertheless, in this paper we will use the terms introduced in Langacker’s (1987) Cognitive Grammar, trajector and landmark. These notions denote perceptual features that are pivotal to the conceptualisation of spatial relations. The trajector is movable and of small size, when compared to the landmark, whose position is stable and larger. The trajector is the entity whose location is specified. On the other hand, the landmark constitutes a very salient reference; it is the background against which to locate the trajector, which receives the higher focus of attention. These two notions can also be paralleled with the figure and ground distinction advanced by Talmy (1978). 3. THE PREPOSITION AT: THE STATE OF AFFAIRS IN PRESENT-DAY ENGLISH The preposition at is an example of what is called in pertinent literature a topological preposition. Topology is a concept from the field of geometry that refers to the relations that are preserved under certain deformations such as bending and stretching. Even though, in principle, topological prepositions can be treated as conceptually simple; there are degrees of complexity within this group, some code axial properties or other intrinsic 2 With the exception of the phenomenon known as preposition stranding: “This is the bed Henry VIII slept in”. The Locative Uses of the Preposition In the OE Version of Bede's EHEP 121 properties of the landmark configuration. Thus Levinson (2002: 72) points out that a topological preposition such as under involves the vertical absolute dimension. This is not the case with at, which can be considered as a prototypical topological preposition; it does not express any information from Euclidean geometry. The topological relation that defines the preposition at is ‘coincidence’. This definition corresponds to Herskovits’ (1986) notion of ideal meaning. As sketched briefly above, the ideal meaning of a preposition is the result of a geometric abstraction of the elements related by the preposition. This geometric abstraction can be inferred from the different senses of the preposition, which Herskovits designates “use-types”. Some of these use-types show that by using the preposition at, the speaker takes a remote perspective of the scene being described. It is precisely this fact that favours the conceptualisation of the trajector and the landmark as points. There exist contrasting pairs that reflect different viewpoints of apparently similiar situation. Consider the following locative expressions: (1) a. Mary is at the shop b. Mary is in the shop The expression 1a. is less specific than 1b. By uttering “Mary is at the shop” the speaker does not fully commit to the location of Mary within the physical boundaries of the shop. Maybe, she is still on her way to the shop, or she could already be on her way back from the shop. The speaker may not even know the exact reference of the shop, and by uttering that expression, she may simply be implying that Mary is engaged in the activity of shopping. We agree with Herskovits (1986) that in examples like this one can clearly predict that neither the speaker nor the interlocutor are in the shop. On the other hand, 1b. would be uttered in a context where the speaker has visual access to the situation described by the expression, or at the least, the speaker and her interlocutor are both in the shop. The use of these two prepositions in these pairs is modulated by the view that the speaker has of a certain scene. This contrast also holds between at and other prepositions such as on, next to, or near and can be expressed in terms of a remote versus a close-up perspective of a scene. The remote view associated with the preposition at is consistent with other conditions that its landmarks must satisfy, i.e. they should not be very large entities. Carmen Guarddon 122 While Herskovits has established a number of use-types for the topological prepositions, offering a polysemic view of the semantics of prepositions, the structuralist Bennet (1975) postulates the existence of a single meaning for the preposition at which is modulated by contextual influences. He argues that the feature that defines all the uses of at is ‘locative’. Therefore, according to Bennet (1975: 66) the at of “arrive at” or “throw at” or “Gwyneth is at the supermarket” are all realisations of the element ‘locative’. One of the factors that distances Bennet from most cognitive linguists is his reluctance to ascribe geometric dimensions to the meanings of prepositions.3 Following this line, Bennet does not concede a relevant role to the speakers’ ability to abstract from the real characteristics of the entities that take part in a spatial relationship nor to their capacity to conceptualise in terms of geometric idealisations: “From a mathematical point of view, the touch-line running the length of a football field is no doubt one-dimensional, but no one has ever seen a one-dimensional touch- line. To be visible a touchline (sic) needs to have width as well as length” (1975: 71). Other linguists agree with Herskovits in associating the relationship ‘coincidence’ with the preposition at, for instance, Boggess (1978). Boggess also points out that at has many of the special contextual conditions of the preposition to, although they are differentiated by the fact that at is not used to express motion. She argues that both prepositions can be used in order to refer to scenarios. Scenarios like 1a. evoke the characteristic that the specific place is not relevant but the connotations associated with it are. Thus, if one says “John is at the doctor’s” or “John must go to the doctor’s”, what is at issue is that John is in a situation that forces him to use the services of a doctor and not his potential location. We contend that there is a further feature which associates to and at. Both prepositions can encode the goal of a trajectory; this is intrinsic to the preposition to because it is a path preposition, and it is a characteristic acquired by the preposition at depending on the compositional sentence processes in which it participates. In other words, we adhere to a current school of thought that holds that 3 Despite not being a cognitive but a structuralist linguist, Leech (1969) pioneered the description of prepositional meaning using geometric features of dimensionality. The Locative Uses of the Preposition In the OE Version of Bede's EHEP meaning is not contained within a single lexical item but is distributed over all the elements that compose a sentence (cf. Levinson 1991). In the particular case of relational elements that do not have direct referents, such as prepositions or verbs, it is their argument structure that is fully responsible for constraining their semantic potential. For instance, the preposition at introduces the end-point of the trajectory when it collocates with projection verbs such as “look” or “throw”. Projection verbs should not be confused with motion verbs; it must be noted that in present-day English the preposition at does not occur in combination with the latter. The following examples can be rated as ungrammatical: (2) a. *He came at home b. *Mary went at the doctor’s Nevertheless, in Old English, even though infrequent, there are examples where the preposition at occurs with motion verbs (cf. Belden 1897; Lindkvist 1950): (3) a. Ic eow cleopode ær to me, ac ge me noldon æt cuman (King Alfred’s West-Saxon Version of Gregory’s Pastoral Care: 247). b. ñæt hie næfre noldon æt ham cuman ær hie òæt gewrecen hæfden (King Alfred’s Orosius: 122). The fact that this colocation is no longer grammatical may be interpreted as evidence of a limitation of the field of spatial relationships that the preposition at covers. This view is supported by the work of Lundskær- Nielsen (1993: 84). We believe that this pattern has been partially inherited by the present-day use of this preposition, as it shows its capacity to occur with verbs that focus on the end-point of a path, such as “arrive at” or the projection verbs mentioned above. However, the evolution of this preposition does not necessarily respond to a limitation of its categorisation field. There were also temporal extensions of the semantic range of the preposition at that did not survive to present-day English. For example, in his classification of the Middle English uses of this preposition, Lindkvist showed that it occurred in examples of location in a country: “An interesting combination of first extension and then limitation of the application of at is found in its use in Middle English and Early Modern English with complements denoting countries and similar large areas” (1978: 27). 123 Carmen Guarddon 124 Furthermore, conventionally allowed deviations from the ideal meaning are a constant source for extensions in the use of the category. The time that these extensions will remain within the language use can be relatively short as Lindkvist example shows, however this is something almost impossible to predict on the ground of the data available to a linguist. Lindkvist (1950) also carried out a classification of the uses of the preposition at in present-day English. He left open the question of whether all spatial uses of a preposition can be accounted for in terms of a single sense. He does not take into consideration questions of abstraction in the conceptualisation of spatial relations. Lindkvist sets up a classification of the uses of at by studying a large corpus formed by the occurrences of this preposition in a huge number of English literary works. His classification includes the following uses: I. Location in close proximity to an object; II. Location within an area or space or on a surface apprehended as a point; III. Relative position; IV. Location close to or within a body, surface or area thought of as being used to serve a certain purpose; V. Motion and direction. Finally, there is a use of the preposition at which cannot be explained without recourse to functionality; such as the cases in which the trajector presents the feature [+human] and the landmark is an artifact that the trajector is using, i.e. “Mary is at her desk, John is at the piano; Eloise is at her computer...” The pivotal question in these expressions is that the relationship of physical coincidence that holds between the trajector and the landmark allows the first to use the second. However, this coincidence is not enough for that relationship to qualify as a case of at. If John happens to be sitting on the upper surface of the piano, the relationship between them would have to be encoded by a preposition other than at – most probably on. The different positions or postures of the trajectors when using a certain artifact must be inferred pragmatically. This use is defined by Herskovits as ‘person using artifact’ (1986: 135) and by Lindkvist as ‘activity close to an object’ (1950: 165). 4. THE PREPOSITION AT IN THE OLD ENGLISH VERSION OF BEDE’S ECCLESIASTICAL HISTORY OF THE ENGLISH PEOPLE The Locative Uses of the Preposition In the OE Version of Bede's EHEP 125 The English preposition at is cognate with the Latin ad “to, at”. Even though, in Old English it governs the dative only, the texts examined by Belden (1897) present the preposition at followed by the accusative in one case, the phrase æt òysne andweardan dæg “at this present day”. One of the main reasons for choosing EH to create our corpus is that being quite long – it is made up of five books – we have found enough occurrences of the preposition at to spot certain tendencies in the use of this preposition. Second, the three main topological prepositions, in, on and at, which have a rather restricted distribution, are well represented in the text. This allows us to establish the selection restrictions governing the distribution of the preposition at in EH, which can also be extrapolated to other Old English texts. In fact, Lundskaer-Nielsen’s (1993) asserts that at presents quite a stable usage over texts from different dialectal areas. This estability is also attested by Lindkvist (1978), who examined the usage of the preposition at in Old English across several texts of this period (Beowulf, The Anglo-Saxon Chronicle, The Old English Version of Bede’s Ecclesiastical History, Alfred’s Translation of Gregory’s Cura Pastoralis, etc.) (cf. Hols 1971). We have examined the behaviour of at in relation to a number of spatial entities. There is a widespread assumption that prepositions are highly polysemous words and thus form extremely complex lexical classes. Their polysemy is based on their capacity to categorise domains other than the spatial. Therefore, the semantic description of a preposition is not complete if no reference is made to the temporal and abstract relationships that it can establish. However, these relations lie beyond the scope of the present paper. At any rate, due to their ontological priviledged status, the spatial uses of a preposition should be described in the first place in order to determine the metaphorical and metonymical operations that have led to further extensions of the category in other domains. The landmarks selected for the purposes of our analysis stand for a wide range of spatial categories: large geographic entities, small geographic entities, general geographic designations, buildings, containers, body parts, means of transport and imaginary places. The table below shows the catalogue of landmarks that have been used to build up our corpus: Table 1. Physical landmarks from EH. Carmen Guarddon 126 1. Briten “Britain” 8. Gallia “Gaul” 15. Mynster “monastery” 2. Burg “town, city” 9. Hand “hand” 16.Neorxenawang “paradise” 3. Cirice “church” 10. Heofon “heaven” 17. Ríce “kingdom” 4. Cyst “coffin, 11. Hús “house” 18. Rom “Rome” 5. Eòel “country, 12. Land “land, country” 19. Scíp “ship” 6. Fót “foot” 13. Mægð “province, tribe” 20. Stow “spot, place” 7. Fyr “fire” 14. Mór “moors” 21. Ðruh “coffin” It must be noted that with some of these landmarks we have found no examples using the preposition at, but they have been kept because, we claim, that for the semantic description of a preposition, indicating the type of landmark it collocates with is as relevant as enumerating the type of relationships that it cannot establish. At this point, it must be noted that EH is the translation of a Latin original and one could argue about the possibility of Latin influence in the prepositional usage of the Old English translation. To check whether such dependence has biased the use of at, we have compared all the expressions that compose our corpus with their Latin equivalents and we have not found any repeated synchronicity in this sense, for example, the Latin preposition ad motivating the presence of at in the Old English text. In his study of the case values governed by prepositions in Old English Belden (1897) also confirmed that the Latin work had not determined the choice of prepositions in EH. In the next section, we show the usage potential of the preposition at in the physical domain, and how this potential can be subsumed under a classification of use-types. Our main interest is to establish the extent to which these use-types coincide with the present-day usage of this preposition. 4.1 EXAMINATION OF CORPUS AND RESULTS Unless special contextual circumstances co-occur, for instance the landmark is part of the case frame of a certain verb, the entities that belong to the same type of spatial category should collocate with the same topological preposition. This is the state of affairs in present-day English; The Locative Uses of the Preposition In the OE Version of Bede's EHEP 127 differences in this sense should point at the diachronic evolution of the preposition at. As expected, no examples of the preposition at have been found with large geographic entities (Briten, Eðel, Gallia, Mægð, Ríce), even though, as noted above, Lindkvist (1978) provides evidence that demonstrates that in Middle English at was used to express location in a country. However, this was only a temporal extension of the usage of the preposition. The next spatial category analysed is ‘small geographic entities’ such as burg, and Rom. The location in burg is not expressed in a uniform way; the three topological prepositions, in, on and at, precede this landmark.4 There are 12 expressions that use the preposition in, 3 with on and 2 with at. Concerning Rom we find a similar situation. There are 6 locative expressions where Rom occurs with the prepositions in question; there are 3 cases with the preposition in, one with on, and 2 with at. After checking that this distribution does not respond to restrictions imposed by the subcategorisation frames of the verbal predicates of the sentences, we have examined the content of these expressions. It is necessary, first of all, to consider how the preposition at functions in present-day English. As argued in Herskovist (1986), the preposition at in present-day English only occurs with small towns. This is consistent with the two expressions in which at collocates with burg (1, 2). The landmark in both of them is Coldingham; most probably this town was not at the time as well known to the Britons as other cities mentioned in EH, such as Canterbury, London or Rome. One of those expressions (2) presents the particularity that the preposition has become part of the name of the town. The prepositional phrase serves to refer to the name of a monastery, which is At Coldingham. This indicates the strong association existing between a trajector and the place where it is situated: ðæt mynster, þæt mon nemneð Æt Coludes burg (2). According to Lindkvist (1978: 12), the preposition at should not be translated in these cases, since the whole prepositional phrase is understood as a place name. He also points out that these prepositional phrases function as a predicate instead of as an adverbial or an adjunct. 4 The examples with the preposition at are listed in an Appendix at the end of this article. When referred to or partially quoted, these examples are identified with a number that corresponds to the sentences listed in the Appendix. Miller’s translations are also provided. Carmen Guarddon 128 The explanation provided above for the collocation of at and burg falls in contradiction with the two expressions where this preposition occurs with Rom. In fact, Rom was already at the time of the translation a large and important city. As nothing in the semantic content of the sentence where the first of these expressions (3) appears sheds any light on the factors motivating the presence of at, we have looked at its broader context. We could observe that this case occurred when the author steps off the narrative line and digresses in order to relate the biography of a specific character. In biographies, the relevant aspects are the character’s deeds and achievements, dates and places usually play a lesser role. This factor motivates a minor involvement of the speaker with the locative expressions at issue, which is consistent with a remote perspective.5 Regarding the second example (4), the reasons for the occurrence of the preposition at were found in the neighbouring context. A sequence of events, the consecration of two bishops, is enumerated with the different places where the events happen, one of those places is Briten, preceded by the preposition on, and the second is Rom, preceded by the preposition at. A mental map of the places is needed to account for the different conceptualisations of these spatial entities. The occurrence of these two prepositions serves to emphasise the smaller size of Rom in comparison to Briten: Ðæt se arwurða wer Swiðbyrht on Breotone, Wilbrord æt Rome biscopas wæron Fresna ðeode gehalgode “That the venerable Swithberht in Britain and Wilbrord at Rome were consecrated as bishops for Friesland” (EH: 22). It must be noted that Lundskær-Nielsen (1993) found out that the codification of location in a town is one of the most frequent locative uses of at in the section of the Anglo Saxon Chronicle that covers the years 892-900: æt Middeltune, æt Beamfleote, æt Cwatbrycge, etc. Interestingly enough, in this section there are three instances of towns that collocate with the preposition on and these are remarkably more important and larger than the ones mentioned above: Lundenne, Hrofesceastre and Wintenceastre. In fact, these three cities had been endowed with a bishopric in that time. 5 We would like to thank one of our anonymous reviewers for noticing that the use of this preposition with larger landmarks in Old English can anticipate this use of at in Middle English. In fact, Lindkvist has found several Middle English cases in which this preposition collocates with important cities: “The heved is at Parys. Maundev.2.; He was at Jerusalem in tho dayes. Wycl. Lucke 23. 7.” (1978: 13). The Locative Uses of the Preposition In the OE Version of Bede's EHEP 129 The third category we have set up in relation with the landmarks from Table 1 is ‘general geographic designations’. Under this category we include places that vary in size and are not conceptualised as bounded. In our corpus there are three: land, stow and mór. Land occurs 23 times in EH, in 15 cases it occurs with the preposition in; there are 7 examples with the preposition on, and, finally, only one with the preposition at. In principle, land refers to an area, therefore it may seem difficult to explain how an extension in space can be conceptualised as a point. However, the only case where at precedes land is inserted in a very specific context: ðurh ðone smyltestan sae usic æt londe gebrohte (5). Here land does not refer to an extension, instead it is part of the opposition sea versus land. Thus when one speaks of travellers reaching land after a long journey in a ship, “at land” is said to indicate the point where the travellers arrive in a different medium. It can also be viewed as the goal of a path through the sea; this idea is consistent with the consideration of the landmarks of at as points. Lindkvist (1950: 134) pointed out that the preposition at occurred with the noun land in Early Modern English and even later “to indicate situation on the solid surface of the earth in contrast to the sea”. But as we can see this usage of at has antecedents in Old English. With regard to stow, it is even more general than land; this flexibility may account for the large number of occurrences of this noun in EH. We have found 74 expressions with the preposition in; 18 with the preposition on, and 3 with the preposition at. The expressions occurring with the preposition at are those that refer to “the holy places of the apostles” (6, 7, 8). From the context, we learn that those places are situated in Rome; this helps to establish a remote perspective on the landmarks as they are not even situated in Britain. There is one more factor motivating a remote perspective of these places: there is no information about them, we do not know whether they are buildings, or other types of location: þæt he wolde to Rome feran þær æt þam halgum stowum his lif geendigan (6). The third geographic entity examined, mór, resembles the former in the sense that it may refer to an extension of variable size and it does not have salient boundaries. The location specified in the two expressions found in EH is effected by the preposition in. The presence of in in the two examples reflect the close view of the narrator, as the landmark is preceded by an adjective describing the Carmen Guarddon 130 moors internally and the facts that take place there are known, i.e. in heaum morum “high up upon the moors” (EH: 364). The fourth category is ‘buildings’. In our corpus we have included cirice, hús, and mynster. We support Herskovit’s (1986) and Bennet’s (1975) semantic treatments, which claim that the preposition at does not profile any part or subdivision of the landmarks. Bennet (1975: 68) states this fact explicitly: “Whereas on and in are ascribed a ‘part’ component, no such component is attributed to the meaning of at.” As noted above, Herskovits defined the categorisation labour effected by at as ‘coincidence’. We claim that the conceptualisation of physical location in buildings as coincidence is in principle improbable, because buildings are not likely to be idealised geometrically as a point. Human beings have an active bodily and visual interaction with buildings. We are aware of being enclosed by their walls and ceilings, and are able to have a full visual picture of their exterior. However, the relation between the trajector and the building may be not so much one of location, but, instead, the speaker implies that the trajector is engaged in the activities normally performed in that building. In this case, in present-day English, the preposition at is preferred with a large number of landmarks. Concerning the examples analysed here, we have not found any case of collocation of the preposition at with any of the buildings considered, despite the fact that these landmarks are notably represented in EH; there are 68 examples of cirice, 16 examples of hús and 76 of mynster being preceded by the preposition in and on. This seems logical in the case of hús because no activities are specifically associated with this landmark. On the other hand, with cirice and mynster we expected to find some examples with at. In fact, Lindkvist’s (1978) findings show that Old-English at was used to express somebody’s location in connection with the activities typically developed in a building. For example, in present-day English church may be preceded by the preposition at when the church is not referred to as a specific physical entity, i.e. a building, but as the institution where religious activities are held. It must be noted that in these cases the noun is not preceded by any modifier, which emphasises that it does not have a specific referent. In our corpus, there are expressions where church is not accompanied by any modifier either, thus we may infer that the institution and not the building is referred to. The Locative Uses of the Preposition In the OE Version of Bede's EHEP 131 We have examined these examples more closely in order to check whether any other factor was preventing the occurrence of at in these prepositional phrases. In fact, a close examination of the cases where cirice is mentioned as an institution shows that many of them describe the circumstances under which a person should not enter a church, because trespassing on its grounds would be regarded as a sin, for example menstruating women: ne sceal heo bewered beon þæt heo mote in circan gongan “it is not right that she should be cut off from entering God’s church” (EH: 78). Thus, even though not a specific church, but a space clearly delineated is being referred to, enclosure within that space becomes salient, since being in or keeping ourselves out determine whether one has committed a sin. The salience acquired by the physical limits of the landmark motivate the presence of the preposition in in these expressions. Finally, as an additional factor to be taken into account, it must be noted that the translators are clerics, which means that they are very familiar with the interior and the activities developed in places such as a church or a monastery. This fact prevents the translators from having a remote perspective of these locations. The following category is ‘containers’. In this respect we have analysed the spatial expressions that occur with cyst and ðruh. The entities within this category belong to the class of objects that have an inherent interior. Consistently with the state of affairs in present-day English, no cases with the preposition at have been found to express containment in relation to these landmarks. In the unlikely case the preposition at appeared with these objects, it would express spatial coincidence, or proximity. In present-day English it is frequent to see parts of the body preceded by the preposition at. Most of these expressions have a figurative meaning (“I will always be at your feet”) but we contend that they started as literal expressions. Some of the examples found in EH confirm this hypothesis. There are 15 locative expressions with “hand”, 3 collocate with the preposition at. Two of them present a literal meaning of “at hand” (9, 10); they refer to an individual involved in the story who does not have something he needs at that moment, a possession that is usually wielded in one’s hands. For instance: næfde he scyld æt honda (10), where one of the king’s attendants cannot find a shield. This type of expression may be at the Carmen Guarddon 132 origin of those already used in Old English and still in present-day English, where the prepositional phrase “at hand” refers to something that is not available, it does not matter if it is something that can be held or not: “Sorry, I cannot pick you up now, I have no car at hand.” In the third example (11), in contrast with the other two, the hands are preceded by a noun in the genitive case, this modifies significantly the meaning of the prepositional phrase. It no longer refers to whether something is or not easily accessible. By mentioning specifically the person whose hands are being referred to by the landmarks, the actor of an action is alluded. This new semantic content of the locative expression may be motivated by the fact that the hands are the part of our body with which we act most often. The prepositional phrase in question, æt þaes biscopes honda, refers to the bishop that gives holy communion to one of the characters of the narration. We claim that this expression also highlights the proximity that there must obtain between the agent and the patient, which is also connected with the postulated coincidence between cause and effect. Wood argues that this use can be identified in present-day English in expressions such as “The captive suffered much harsh treatment at the hands of his jailers” (1967: 285). The next body part we have examined is fót, there is only one case preceded by the preposition at (12). This example is not part of a figurative expression of admiration or respect, the feet in question belong to an individual lying in bed while another person is standing close to the foot of the bed: oðer æt minum fotum (12). In conformity with the meaning of the preposition at, although the spatial relation is actually of proximity, it is conceived of as coincidence. The literature on English prepositions generally define one use of the preposition at as being “activity close to an object”. The trajector presents in most cases the feature [+ human] while the landmark is an object with which the trajector interacts. This interaction requires the physical coincidence between them. In EH we have examined the locative expressions containing fyr and the preposition at. One must be close to, or be physically coincident with, the fire in order to get warmer or cook food. The relevant fact in such a relationship is the physical coincidence between both trajector and landmark. Another reason for preferring the preposition at in this case is that no part of the landmark is profiled as being the exact location of the trajector. There are The Locative Uses of the Preposition In the OE Version of Bede's EHEP 133 only two prepositional phrases where fyr is the landmark of one of the topological prepositions; one of the examples presents on and the other presents at (13). This expression, as expected, includes the sense of functionality that we have mentioned above: gestód æt þam fýre hine wyrmde mid his þegnum. As this sentence shows, the purpose of the trajector by standing close to the fire is to warm himself. This usage has not changed over time and it is highly productive in present-day English, as Lindkvist put it: “At is used with complements denoting different things which according to their nature are the objects or centres of special activitites, to indicate that the activity connected with such a thing is carried on close to the thing” (1950: 165). When it comes to means of transport, we have examined the prepositional phrases with the word scip and have found 5 cases with the preposition in and 3 with the preposition on; there is not a single occurrence of this word preceded by the preposition at. The reasons for this absence can be attributed to the way human beings interact with the means of transport that resemble a container, such as scip. Sometimes, it is quite similar to the way we interact with prototypical closures such as buildings in the sense that we are aware of being situated within their limits; this type of location motivates the use of the preposition in. On the other hand, in other cases, the horizontal surface that supports the travellers is the part of the landmark that obtains a maximum salience from the cognitive point of view, this relationship is encoded by the preposition on. In contrast, the preposition at would not profile the interaction of the traveller with any of the parts of the means of transport. Finally, we have examined locative categories that human beings have not experienced physically, places such as heofon and neorxenawang. We have a culturally-built mental image of these places, and of what one can find in their interior. This mental image is commonly shared due to the contribution of art, specifically painting, and descriptions in religious writings i.e. The Phoenix. Before the search, we advanced the hypothesis that the relatively deep knowledge that scribes had of what the interior of these regions should be like precluded the occurrence of at with these landmarks. Furthermore, their exact situation does not rely on the principles that human beings resort to when locating, i.e. distance, the points of the Carmen Guarddon 134 compass, and map conventions in general. Therefore, speakers cannot be said to have a remote perception of its position in space either, which is one of the main features of the preposition at. This hypothesis is confirmed after examining the data; there are 4 locative expressions with heofon, 3 with the preposition in and one with the preposition on. Another text highlights the scarce possibilities of finding the preposition at preceding landmarks of this nature. For instance, we have also examined Satan in Hell, inserted in the section of Genesis known as Genesis B (Vickrey 1960). There heofon appears on 7 occasions and in all of them is preceded by the preposition on. In the same text, the word hell, which belongs to the same category of landmark, occurs 6 times, all of them preceded by the preposition on. It must be noted that the preposition at appears in this text in combination with other landmarks. Regarding neorxenawang, it occurs just once preceded by a topological preposition, which as we predicted is not at but in in EH. This word occurs twice in Ælfric’s fragment from the Book of Genesis known as The Fall of Man (Mitchell and Robinson 1992), in both cases it is preceded by the preposition on. It is worth noticing that in the same text we have found the Latin word ParadĄsum, also preceded by the preposition on. 5. DISCUSSION OF RESULTS The question we want to address in this section may be formulated like this: Does the allegedly ideal meaning of at apply to all the uses of the preposition that have been attested in our analysis? In the first place, the examination of the landmarks selected with the preposition at in this Anglo- Saxon document shows that, in contrast with in and on, the usage of the preposition at in this period is highly consistent with present-day English. Therefore, it could be claimed that the ideal meaning proposed for this preposition according to its usage in present-day English, ‘location at a point’, applies in general to the way it is used in Old English. This is supported by Lindkvist’s claim that “With very few exceptions […] the use of at in Middle English is identical with that of æt in Old English” (1978: 13). This indicates the existence of a continuity in the usage of this preposition. But this is not a strong position; the question is whether ‘location at a point’ is a suitable definition for the totality of the senses of at. We contend that map conventions to a large extent resemble the way we The Locative Uses of the Preposition In the OE Version of Bede's EHEP 135 think of spatial entities and this is shown by the data examined. For instance, no case of location in a country has been found with at in EH. In fact, large entities tend to be excluded from the set of likely landmarks of this preposition because they are hard to conceptualise as points. With smaller geographical entities such as towns, the expressions with at in EH are appropriate under special contextual conditions. Maps are also flexible concerning the representation of towns, depending on the scale, they can be represented as areas or points. Herskovits (1986) also pointed out that in present-day English the size or the importance of the city determines the distribution of in and at. In our corpus, only small distant towns occur with the preposition at, with the exception of Rom, which is linked on two occasions with this preposition. But in the two cases the location in Rom appears as secondary information, no close-up view of the town is taken and the knowledge of the position of the facts narrated with respect to the landmark is rather imprecise. Thus, these examples show that the feature ‘remote perspective’ was already probably active in Old English in determining the distribution of this preposition. With generic geographic designations, it is possible for the landmark to be conceptualised as a point, which controls the use of at; otherwise, any other preposition, in or on, should precede the landmark. For example, as noted above, land, infrequently, can be conceptualised as a point. This indicates that the determinant factor to decide whether an expression with a preposition is appropriate is not simply the type of spatial entity it occurs with but the contextual conditions as well. Diachronically, contextual conditions as a factor affecting the behaviour of prepositions has declined, although it is still active (consider “at the market” versus “in the market”; “at the bridge” versus “on the bridge”). It seems obvious that the distribution of these prepositions in present-day English is more constrained by the spatial properties of the landmark. Regarding the occurrence of prepositions with buildings, two different situations may hold. On the one hand, locations in buildings usually profile their interior, but the preposition at implies a global conceptualisation of the landmarks. When a speaker uses at she does not commit to the exact position of the trajector with respect to the landmark, whereas by expressing location in a building most often a relation of enclosure is stated. On the Carmen Guarddon 136 other hand, the noun may refer to a scenario rather than to a building. This use has on the whole remained unchanged since Old English. The data examined shows examples where cirice was not referred to as a specific building, but as an institution; in these examples the noun is usually not preceded by any modifier. Even though our data does not include any case of collocation of at with a building in reference to a scenario, Lindkvist provides examples from various Old English literary works that prove that this use was already consolidated in Old English (1978: 13-4). Again, in this case, contextual dependencies determine the selection of a preposition and the rejection of another. These contextual dependencies are responsible for the way a particular relation is conceptualised. In the previous section, we observed the absence of at with cirice when location in a particular scenario was being expressed in EH. So the presence or absence of at in Old English was not entirely predictable with buildings, when denoting the trajector’s engagement in a specific activity connected with the building. Perhaps this instability is responsible for some alternations of the prepositions at and in in present-day English with no difference in meaning: “in/at church” (Swan 1995). The examples examined above mainly show that when the landmark cannot be idealised into a point, because its interior plays a major role in the locative expression, at is not a likely option. The salience of the interior of a landmark is also associated with visual perception. When the speakers have visual information of the interior of the landmark, other prepositions rather than at are used, instead in is preferred if a relationship of enclosure holds. This was the situation of the examples in the corpus that express location in imaginary places like heofon or neorxenawang. The fact that these landmarks have just one single referent, i.e. in the Christian tradition there are not several “heavens” or “paradises”, prevents that a sense of indefiniteness may be involved in these landmarks, which is another factor accompanying the preposition at. Visual perceptual access to a scene or situation is a relevant factor modulating the presence of at versus in, since visual perception allows the speaker to verify that enclosure within a space actually takes place. With landmarks that sometimes are not apprehended visually, such as some buildings or geographic entities, the preposition at then may be used. The Locative Uses of the Preposition In the OE Version of Bede's EHEP 137 However, human beings have a recurrent visual and physical interaction with containers; this is reinforced by the action of physically placing objects into them.6 This precludes the preposition at from occurring with nouns denoting small containers and it is also an explanation for the absence of the preposition at with means of transport that contain people. From the uses of the preposition at proposed above it seems quite straightforward that the characteristic ‘coincidence’ that most linguists ascribe to the category at has to be defined in relation to one basic modality in conceptualising space: map images. For instance, the location in small geographic entities of which we have a remote perspective is conceptualised through an idealised geometric conceptualisation of both trajector and landmark. But this fact does not apply to the cases that are conceptualised from the information obtained through a different modality: visual perception. We mean the situations where the spatial coincidence is perceived visually, for example when the trajector, usually human, performs some function associated with the landmark. Herskovits (1986: 82) claims that a spatial relation like “Maggie is at her desk” is viewed as coincidence between two points. But it must be noted that the selection restrictions that determine the correct use of at in such a context determine the distribution of at versus near or of at versus close to in the same context. We believe these selection restrictions are too specific to be abstracted away in a geometric conceptualisation where the trajector as well as the landmarks are viewed as points. There are two examples in the data where the relation between the trajector and the landmark responds to specificities that are apprehended visually and that the listener must reconstruct through a visual image in order to make sense of it. For instance, oðer æt minum fotum (12), where the position of a person with respect to somebody’s bed is stated. This does not simply imply a relation of coincidence, there are other relationships involved here that are apprehended visually and that are recovered in the codification and decodification of the spatial relationship connected with this scene; as an example, the form of beds and the way we interact with them. A similar 6 For a thorough description of how the action of introducing things into our own mouth when we are babies ingrains the notion of containment by deeply entrenching it into our minds, while obtaining a pre-conceptual status see Lakoff & Johnson (1999). Carmen Guarddon 138 interpretation can be made of the expression with fyr, gestód æt þam fýre hine wyrmde mid his þegnum (13); when somebody is said to be located at the fire, he is assumed to be there for a certain purpose, usually for warmth. In order to suit this purpose, the trajector must fulfill certain conditions: he should be placed at a certain distance, neither too close nor too far, he usually does not show his back to the fire, etc. Probably this functional use of at was already active long before the time the extant documents were written. In fact, Lindkvist attests this specific use in different texts, which led him to affirm: “Æt further occurred in Old English in such cases of practical connection where somebody is present close to, or in contact with, an object which is used for some purpose connected with the object” (1978: 18). Thus we cannot have evidence of the contexts where this use emerged. Neither can we know whether these expressions entered the language before those derived from mental maps. However, one thing seems to be clear: the treatment of the expressions with at where functional factors concur cannot be reduced to the statement of mere coincidence between two points. 6. CONCLUDING REMARKS The advent of cognitive linguistics with its emphasis on the speakers’ fundamental role in the organisation of spatial relations has brought the study of prepositions into a new light. Factors that were not even considered within the standard linguistic machinery have acquired an active role in the description of the semantic content of these lexical items; for instance, speakers’ perception, their interaction with space and subsequent conceptualisation of the spatial relations. This means that when it comes to a diachronic study of prepositions dialectal factors or stylistic variance depending on the author are to be considered, but they are not the only conditions when it comes to explaining apparently random uses. The study we have carried out of the preposition at in EH and its comparison with Lindkvist (1978) and Lundskær-Nielsen’s (1993) analyses of the same preposition provides us with a quite comprehensive picture of the usage of this preposition in Old English. Some extrapolations can be made on the basis of the findings exposed here. For example, the use of this preposition has remained quite stable over time and no major changes have taken place, or they have been temporal, as the use of at to express location The Locative Uses of the Preposition In the OE Version of Bede's EHEP 139 in a country, as put forward by Lindkvist. It seems that the system of topological prepositions in Old English was more flexible in order to encode the speakers’ point of view of a locative relations, as the cases where Rom was preceded by at show. For instance, in present-day English, the perspective that the speaker takes of a certain scene is not echoed in the selection restrictions of some prepositions, i.e. at cannot occur with terms denoting large important cities. We can see here two distinct parameters in a continuum competing to determine the use of prepositions: on the one hand the speakers’ perspective of a situation and, on the other hand, the physical objective characteristics of the landmark. The freedom speakers have to choose a preposition according to the perspective they take of a scene is directly proportional to the flexibility of some prepositions to alternate with specific landmarks. In present-day English, this flexibility has diminished due to the fact that more strict collocations between prepositions and landmarks occur. For this reason, the preposition at is more constrained by the type of landmarks it collocates with. This was maybe intuitively realised by Lundskaer-Nielsen when he stated “that later changes in the use of æt are more often limitations in the semantic range that it had in OE than extensions of it” (1993: 84). We, cognitive linguists, in our attempt to demonstrate that even the lowliest grammatical morpheme is meaningful, usually posit multiple senses for these items and treat them as complex lexical categories. The reaction to this situation is to establish a single sense or core from which all the other senses of the category originated, this core sense is called by Herskovits (1986) “ideal meaning.” The French linguist Vandeloise (1984) introduces diachronic aspects in this core sense that he calls “impulsion.” According to Vandeloise, the impulsion is diachronically the primary sense of the category, and linguists should aim at establishing it. Vandeloise makes the point that arguing for the polysemic character of a preposition is a relatively straight point, the hard part comes when ones tries to set up the impulsion of the category. There is no doubt that the notion of impulsion has a high heuristic value. However, concerning a synchronic description of a preposition, we are inclined to regard the establishment of a single definition that suits all its uses, in some cases, as being too artificial. Therefore, as we have argued in the previous section some relationships are naturally Carmen Guarddon 140 conceptualised as geometric idealisations, as our familiarity with the use of maps shows. On the other hand, other relations are apprehended visually and understood in a context where several factors converge, for instance, functional landmarks. The data analysed shows that in Old English as well as in present-day English a single definition is not able to cover all the attested senses of at. Finally, taking the results of our study as a starting point, we envisage two directions for future research. First, we have strictly focused on spatial uses but it would be interesting to verify that the stability of the categorisation of at between Old and present-day English in the spatial domain can be attested in other non-spatial domains, for example time. Second, the same spatial categories proposed here can be analysed in Middle English texts to determine the specific variations that have taken place over these two periods. Carmen Guarddon Anelo Universidad Nacional de Educación a Distancia APPENDIX: OCCURRENCES OF THE PREPOSITION AT IN OLD ENGLISH VERSION OF BEDE’S ECCLESIASTICAL HISTORY OF THE ENGLISH PEOPLE WITH THE LANDMARKS SELECTED (1) BedeHead 4 20.23 XXIIII. Hwilc gesihð sumum Godes were ætywde, ær þam þe þæt mynster æt Coludes byrig mid bryne fornumen waere. XXIV. The vision which appeared to a man of God, before the monastery at Coldingham was destroyed by fire. (2) 4 26.348.27 Ðissum tidum ðæt mynster, þæt mon nemneð Æt Coludes burg, þæs we beforan gemyndgodon, þurh ungemænne synne fyre lege wæs fornumen. The Locative Uses of the Preposition In the OE Version of Bede's EHEP 141 At this time the monastery called Coldingham, already mentioned, was destroyed by fire and flame, through the sin of carelessness. (3) BedeHead 5 22.32 XI. Ðæt se arwurða wer Swiðbyrht on Breotone, Wilbrord æt Rome biscopas wæron Fresna ðeode gehalgode. XI. That the venerable Swithberht in Britain and Wilbrord at Rome were consecrated as bishops for Friesland. (4) 5 8.406.18 Ða wæs ðy nehstan geare, æfter ðam ðe Ceadwala æt Rome forðfered wæs, þætte ðære eadgan gemynde þeodor ercebiscop ald dagena full forðfered wæs, Then next year, after the death of Ceadwalla at Rome, archbishop Theodore of blessed memory died, old and full days, that is, when he was eighty-eight years of age. (5) 5 1.386.11 Ond mid ðy he ða ðæt gebed gefylde, he þa somod ætgædre ge þone aðundnan sae gesmylte ge ðone storm gestilde, to þon ðætte þurh all sio roeðnis ðæs stormes wæs blinnende gesyndge windas ðurh ðone smyltestan sae usic æt londe gebrohte. And when he had ended the prayer, he then at the same time calmed the swollen sea and stilled the storm, so that altogether the fury of the storm ceased and favouring breezes carried us to the land over the calmest of seas. (6) 4 5.274.29 þæt gif he from þære untrymnesse gehæled wære, þæt he wolde to Rome feran þær æt þam halgum stowum his lif geendigan, ond Wilferð biscop bæd, þæt he him þæs siiðfætes latteow wære, that, if he had been cured of this disorder, he intended to proceed to Rome and there end his life at those holy places, and he begged bishop Wilfrid to be his guide on the journey, Carmen Guarddon 142 (7) 5 7.404.16 þæt he æt stowe þara apostola mid þa wællan fullwihte bæthes athwegen wære, in tham anum he geleornode monna cynne ingong geopenian þæs heofonlican lifes. in the home of the apostles, whereby alone he had learnt that entrance into the kingdom of heaven is opened to mankind. (8) 5 17.448.23 Forðon he com to Rome, þaer scire onfeng munuc wæs geworden on Constantines tidum þaes papa: æt þara apostola stowe on gebedum on fæstenum on ælmesdaedumawunade oð þone ytemestan dæg. For he went to Rome and there received the tonsure, and became a monk in the time of pope Constantine; and continued in the home of the apostles, praying and fasting and giving alms, up to his last day. (9) 2 8.122.19 Ða þæt þa Lilla geseah, se cyninges þegn him se holdesta, næfde he scyld æt honda, Now when Lilla saw this, who was the most devoted of the king’s attendants, having no shield at hand, (10) 3 1.156.27 Ða sæt he æt beode, næfde þa æt honda hwær þæt brohte lac gehealdan scolde; As he sat at table, he had nothing at hand to keep the proffered gift in; (11) 3 5.168.5 Ða onfeng he him nom æt fulwihte bæðe æt þaes biscopes honda þære godcundan þegnunge him to godsuna. and he received and took him as godson at the font, after the sacred administration by the bishop’s hands. The Locative Uses of the Preposition In the OE Version of Bede's EHEP 143 (12) 5 14.438.21 Ær hwene ðu come, eode inn on þis hus to me twegen geonge men fægre beorhte, gesæton æt me, oðer æt minum heafde, oðer æt minum fotum. A little before you came, there came in to me in this house two young men fair and bright, and sat down by me, one at my head, the other at my feet. (13) 3 12.196.25 Se cyning þonne, forðon he of huntað cwom, gestód æt þam fýre hine wyrmde mid his þegnum. Then the king, having just come from hunting, stood at the fire and warmed himself, among his attendants. REFERENCES Belden, M. 1897: The Prepositions In, On, To, For, Fore, and Æt in Anglo- Saxon Prose. A Study of Case Values in Old English. Friedenwald, Baltimore. Bennet, D. 1975: Spatial and Temporal Uses of English Prepositions – An Essay in Stratificational Semantics. Longman, London. Boggess, L. 1978: Computational Interpretation of English Spatial Prepositions. University of Illinois, Ph.D. Dissertation. Brøndal, V. 1950: Theorie des Prepositions. Introduction à une Sémantique Rationnelle. Ejnar Munksgaard, Copenhague. Cassidy F. G. & R. Ringler 1981: Bright’s Old English Grammar and Reader. Harcourt Brace Jovanovich College Publishers, Toronto. Herskovits, A. 1986: Language and Spatial Cognition. An Interdisciplinary Study of the Prepositions in English. Cambridge University Press, Cambridge. Carmen Guarddon 144 Hols, E. J. 1971: Grammatical Roles of Three Sets of Prepositions in Beowulf and King Horn. University of Iowa, Ph.D. Dissertation. Johnson, M. 1987: The Body in the Mind. The Bodily Basis of Meaning, Imagination, and Reason. University of Chicago Press, Chicago. Lakoff, G. & M. Johnson 1980: Metaphors We Live By. University of Chicago Press, Chicago. Lakoff, G. & M. Johnson 1999: Philosophy in the Flesh. Basic Books, New York. Langacker, R. 1987: Foundations of a Cognitive Grammar. Vol. 1, Theoretical Prerequisites. Stanford University Press, Stanford. Leech, G. 1969: Towards a Semantic Description of English. Longman, London. Levinson, S. 1991: Relativity in Spatial Conception and Description. Cognitive Anthropology Research Group at the Max-Planck Institute for Psycholinguistics Working Paper 1, Nijmegen. Levinson, S. 2002: Space in Language and Cognition. Cambridge University Press, Cambridge. Lindkvist, K. 1950: Studies on the Local Sense of the Prepositions in, at, on, and to in Modern English. Munksgaard, Lund. Lindkvist, K. 1978: At versus On, In, By: On the Early History of Spatial At and Certain Primary Ideas Distinguishing At from On, In, By. Almquist & Wiksell International, Stockholm. Lundskær-Nielsen, T. 1993: Prepositions in Old and Middle English. A Study of Prepositional Syntax and the Semantics of At, In and On in some Old and Middle English Texts. Odense University Press, Odense. Miller, T. ed. 1990[1890]: The Old English Version of Bede’s Ecclesiastical History of The English People. Kraus Reprint, New York. O’Keefe J., & L. Nadel. 1978: The Hippocampus as a Cognitive Map. Clarendon Press, Oxford. Swan, M. 1995: Practical English Usage. Oxford University Press, Oxford. The Locative Uses of the Preposition In the OE Version of Bede's EHEP 145 Sweet, H. 1871-2: King Alfred’s West-Saxon Version of Gregory’s Pastoral Care. [E.E.T.S.], London. Sweet, H. 1883: King’s Alfred’s Orosius. [E.E.T.S.], London. Talmy, L. 1978: Figure and Ground in Complex Sentences. In Universals of Human Language. Vol. 4, J. H. Greenberg, ed., Syntax, 625-649. Stanford University Press, Stanford. Vandeloise, C. 1984: Description of Space in French. University of California at San Diego, Ph.D. Dissertation. Vickrey, F. 1960: Genesis B: A New Analysis and Edition. Indiana. University, Ph.D. Dissertation. Wood, F. 1967: English Prepositional Idioms. St Martin’s Press, New York. * † *