SELIM05.pdf Carole P. Biggam, Selim 5 (1996): 109—115 OLD ENGLISH ? ERU AND MODERN ENGLISH THARF-CAKES In a recent article in Notes and Queries,1 Andrew Breeze tackles the mystery word peru which occurs in a list of commodities which the king was entitled to receive from an estate at Westbury -on-Trym, formerly in Gloucestershire. The source text is a grant of the reversion of land at Westbury and Henbury to the Church of Worcester, granted by King Offa of Mercia at a date of 793- 6.2 By presenting evidence from similar lists in the Welsh and Breton records, Breeze convincingly argues that peru means ‘loaves’. I propose to make a further suggestion in support of this argument, namely, that peru is a cognate of OE peorf, ME therf and ModE tharf, adjec- tives meaning ‘unleavened’.3 These adjectives are cognate with ON pjarfr ‘unleavened, insipid’ and OHG derp ‘unleavened’. Significantly for the fol- lowing discussion, OE peorf also occurs as a noun meaning ‘unleavened bread’. Also relevant are OE Íeorfling, ME therfling-bred and ModE ther- fling, tharfling ‘unleavened bread or loaf’, as well as ModE tharf-cake ‘a cake of unleavened bread; now spec. a flat circular cake of oat-, rye-, or barley-meal, unleavened, and sometimes flavoured with butter and treacle; in the latter case = PARKIN’. Îeorfling is first attested in c. 1050, and tharf-cake dates from the fourteenth century. 1 Andrew Breeze, Old English ? eru ‘Loaves’ in a Westbury Charter of 793-796. Notes and Queries, ccxl (New Series, xlii), no. 1 (1995), 13-14. 2 P. H. Sawyer, Anglo-Saxon Charters (London, 1968), no. 146. 3 Old English definitions will be taken from Joseph Bosworth and T. Northcote Toller, An Anglo-Saxon Dictionary (Oxford, 1898), Middle English definitions are from Hans Kurath and Sherman M. Kuhn, ed., Middle English Dictionary (Ann Arbor, 1956-), and Modern English definitions are from J. A. Simpson and E. S. C. Weiner, ed., The Oxford English Dictionary, 2nd ed., 20 vols (Oxford, 1989), hereafter referred to as OED. Carole P. Biggam ____________________________________________________________________ 114 The obvious difficulty in connecting the Old English forms peru and, peorf is the apparent loss of -f and this problem must now be addressed.1 ? eru occurs in the phrase, Vl lang pero ‘six long pero’, in a list of commodi- ties due to the king.2 The context makes it clear that the mystery word must be in the plural, and it is to be interpreted, therefore, as a neuter noun with short stem. 3 If peru had a paradigm like scip, which would be typical of short - stemmed neuter nouns, it would have a nominative singular form * per, but, if it had the rare paradigm demonstrated by searo, for example, it would have a nominative singular in *pero or *peru. If my suggestion, based an Breeze’s insight, is correct, the nomi native singular would, indeed, be *per, but after the loss of final -f from the earlier perf. To find an explanation for this loss, I further suggest that OE peorf ‘unleavened’ had a close etymo logical, semantic, and phonological connec- tion with the Old English verb *purfan which expresses need and obligation. The OED describes the loss of final -f in tharf, thar, a verb which, today, sur- vives only in Scots, and which derives from OE *purfan. The loss of final -f occurred in Mid dle English, the process apparently beginning with the second person singular of the present tense which was pearft in both Old and early Middle English. The difficulty of pronouncing three contiguous consonants re sulted in the form peart tu and then per tu, which left the impression in the minds of Middle English speakers that the verb stem was per-.4 The new stem spread by analogy to other parts of the verb, and was, eventually, so well es tablished that thar became confused with dare. The earliest recorded example of final -f loss in tharf, however, is given in the OED as c. 1290, and this is considerably later than the date of peru. The grant in which the word peru appears occurs in three manuscripts, one of which dates to the seventeenth century, and has not been considered here,5 and one of which does not include the list of commodities in which peru oc- 1 I gratefully acknowledge the kindness of James M. Y. Simpson of the Department of English Language, University of Glasgow, in discussing philological points with me. 2 Benjamin Thorpe, Diplomatarium Anglicum aevi Saxonici (London, 1865), p. 40. 3 Dorothy Whitelock, ed., English Historical Documents c 500- 1042, 2nd ed. (London, 1979), p. 507, note 16. 4 Variant spellings resulted in the additional stems of par-, por- and pur-. 5 London, British Library, Cotton Vitellius C ix, fo. 130v. Old English ? eru and Modern English Tharf-Cakes ____________________________________________________________________ 115 curs.1 The sole source for peru is, therefore, a manuscript of the first half of the eleventh century, London, British Library, Cotton Tiberius A xiii, often referred to as Hemming’s cartulary.2 From internal evidence, Ker believes that Tiberius I, the part of the manuscript in which the Westbury grant occurs, dates to ‘not much, if at all, later than 1016’.3 It must now be considered whether it is reasonable to project back the known loss of -f in tharf from c. 1290 to c.1016. The dislike of three contiguous consonants was not limited to Middle En - glish. Campbell includes six types of consonant loss in Old English from such groups, one of which includes the loss of -f, and he writes ‘Groups of three or more consonants were often reduced in OE in pronunciation, though the full form generally continues to be writ ten’.4 It is suggested, then, that the sporadic reduction of tharf, thar recorded for Middle English had also occurred in late Old English.5 Clearly, the reduction was not carried out by all speakers in either Old or Middle English, and both forms are instanced into the nineteenth century.6 For some speakers of Old English, the stem per was, apparently, well-es - tablished, since peru exhibits a short -stem case ending. If the stem had been understood as perf, the nominative and accusative plu rals would have had a zero case ending, that is, perf, unless uncertainty over this word had resulted in an irregular paradigm for some speakers. It is also possible that the reduc- tion of perf indicates extreme familiarity with this word in certain circles, since Zipf’s law shows how, the more frequently used a word is, the more 1 London, British Library, Cotton Nero E i, pt 2, fo. 181v. 2 The Westbury and Henbury grant appears on fos. 48-9. 3 N. R. Ker, ‘Hemming’s Cartulary’, in Studies in Medieval History Presented to Frederick Maurice Powicke, ed. R. W. Hunt, W. A. Pantin and R. W. Southern (Oxford, 1948), p. 49-75, at p. 69. He later gave the date as s. xii, meaning c. 1025. See N. R. Ker, Catalogue of Manuscripts Containing Anglo-Saxon (Oxford, 1957), p. 250. 4 A. Campbell, Old English Grammar, corrected 1st ed. (Oxford, 1962), p. 190-2. The example of -f loss is Íorleas from Íorfleas, at p. 192. 5 I am referring here to the date of the manuscript of the Westbury grant, and avoiding any assump tions about the form this word took in the original late eighth century text. 6 See the OED under tharf-cake. Carole P. Biggam ____________________________________________________________________ 116 susceptible it is to being shortened in speech.1 The word may have been a rarity for some speakers of Old English, but to a scribe working on estate records, a word concerning payments in kind to the king may have been con- stantly on his lips. It is just possible, therefore, that peru represents a pronunciation common among estate managers and record keepers. In his article, Breeze writes ‘… it is likely the six Westbury loaves were of the best quality, for the king’s table’.2 This is certainly a reasonable assump - tion, but how can it be reconciled with the fact that unleavened bread is heavier and more difficult to digest than leavened bread, and that it is usually made from types of cereal grain, such as barley, oats and rye? which are generally considered to be inferior to wheat? Quotes from modern times leave us in no doubt of the general opinion of tharf bread: They never gat owse better than thaaf keahyk ( 1850), and As thodd’n as a tharcake ( 1740).3 The EDD, in fact, shows that the adjective tharf acquired some fairly uncompli- mentary connotations in the context of human personality, for example, ‘lumpish’, ‘reluctant’, ‘forbidding’, ‘unsociable’ and others.4 This unfavourable impression of unleavened bread is not evident in An- glo-Saxo n sources, mainly because of its religious significance. This type of bread had an important religious role among the He brews, and, after their conversion to Christianity, the Anglo -Saxons would have been well aware of this from several passages in the Bible. They would read, for example, in Ex- odus 12: 15-20 that the Hebrews were commanded by God to put all leaven out of their houses and eat only unleavened bread for seven days during the Passover, and, in the New Testament, they would read that Christ wa rned his disciples to ‘Take heed and beware of the leaven of the Pharisees and of the Sadducees’.5 Ælfric shows that this concept of leaven as a metaphorical con- taminant of pure unleavened bread was well understood in Anglo -Saxon England: ? eorfe hlafas we bringaÍ gode to lace, Íonne we buton yfelnysse 1 George Kingsley Zipf, Human Behavior and the Principle of Least Effort: an Introduction to Human Ecology (New York, 1949). 2 Breeze, ‘Old English ? eru’, 14. 3 Joseph Wright, The English Dialect Dictionary, 6 vols. (Oxford, 1898-1905), VI, p. 75. Hereafter referred to as EDD. 4 EDD, VI, 74. 5 Matthew 16: 6. Old English ? eru and Modern English Tharf-Cakes ____________________________________________________________________ 117 beorman, on Íeorfnysse, syfernysse and soÍfæstnysse faraÍ.1 ‘We bring un- leavened loaves to God as an offering, wherefore we depart without the leaven of wickedness, in purity, cleanliness and truth’. Thus the Biblical tradi- tion must have suggested to the Anglo -Saxons that unleavened bread was purer and of higher status than leavened bread. I now return to my earlier suggestion that the semantic connection of peorf, adjective and noun, was much closer to the verb *purfan than is evi- dent from dictionaries of Old English. One semantic feature of *purfan con- cerns obligation, expressed by Bosworth and Toller as ‘to need to do some - thing … where the need is based on grounds of right, fitness, law, morality, etc., to be bound to do something because it is right, etc., … with the idea of compulsion, or where the in evitability of a consequence is expressed … to be obliged, … to owe’.2 This sense of obligation, presumed to be present in peorf explains the use of the word for a type of bread connected with reli- gious observances, and with dues payable to the king from certain estates, and it is easy to understand that such bread, because of its purpose rather than its digestibility, would be a high status commodity. The suggestion that peorf included a semantic element concerned with obligation does not necessarily contradict Heyne’s theory that the Old En - glish, Old Norse and Old High German adjectives indicated flatness.3 ON pjarfr is defined by Zoëga as ‘unleavened (of bread), fresh (of water), insipid, flat’,4 and De Vries declares the etymology uncertain, but suggests the pres - ence of semantic elements concerning rigidity (‘steif, starr’), bitterness of taste (‘bitter, scharf, sauer’) and firm strength (‘wohlernährt, fest’).5 I suggest that the superordinate semantic concept is a sense of compression similar to the effects of a weight restraining a person or object. Unleavened bread is flat, as though it had been compressed, a body of fresh water sits, apparently heavy and unmoving as though compressed, and quite unlike the apparent 1 Malcolm Godden, ed., Ælfric’s Catholic Homilies: the Second Series (London, 1979), p. 120, lines 364-6. 2 Bosworth and Toller, Anglo-Saxon Dictionary, under purfan, meanings II: 2 and 3, and III. 3 Moriz Heyne, Funf Bücher deutscher Hausaltertümer von den ältesten geschichtlichen Zeiten bis zum 16. Jahrhundert: ein Lehrbuch, 3 vols. (Leipzig, 1899-1903). See vol. 2, p. 2689. 4 Geir T. Zoëga, A Concise Dictionary of Old Icelandic (Oxford, 1910). 5 Jan De Vries, Altnordisches Etymologisches Wörterbuch, 3rd ed. (Leiden, 1977).s Carole P. Biggam ____________________________________________________________________ 118 reaching upwards of the waves of the sea. The reconstructed verb *purfan representing a sense of obligation, indicates the weight of unavoidable duty on those who are thus restricted. It is understandable that the abstract sense of the weight of obligation should result in a metaphoric application to a flat- tened and dense object. Once this application was made to unleavened bread, the semantic implications connected with taste would follow. Another semantic feature of *purfan concerns being in need, expressed by Bosworth and Toller as ‘to be in need, have need of some thing’.1 Poverty and need also result in the sense of a weight bearing down on the needy person, and it is common to speak of ‘a pressing need’. This sense may also have been present in peorf when used of the flat breads made by poorer families who did not have access to the preferred wheat flour.2 Such bread, and its terminology, must have had a history reaching from the Saxon habitation of northern Germany to the uncomplimentary modem comments on tharf-cakes quoted above.3 The twin concepts of the need to render dues to high-ranking persons, and being in need oneself, as contained in *purfan resulted in the apparent contradiction of unleavened bread being of high status in Anglo -Saxon England and of low status in later times, although the latter sense may also have been present in the pre -Conquest period. Finally, it is possible that one other example of OE peorf without final -f is extant. In the Old English translation of Leviticus 8: 2, God tells Moses to take Aaron and his sons, and all the people, to the doors of the tabernacle, with animals, ritual clothes, anointing oil, and peorfe hlafas ‘unleavened loaves’.4 At verse thirty-one of the same chapter, after the ritual clothing of Aaron and his sons, and the sacrifice of the animals, Moses tells them to cook the meat and … etaÍ pær pa … Crawford gives the equivalent Latin phrase from the Vulgate text as … ibi comedite eas … and, presumably, regards k as a transla - 1 Bosworth and Toller, Anglo-Saxon Dictionary, under purfan meaning I. 2 Ann Hagen, A Handbook of Anglo-Saxon Food: Processing and Consumption (Pinner, 1992), p. 9. Hagen discusses leavened and unleavened bread on p. 7-9. See also Ann Hagen, Second Handbook of Anglo-Saxon Food and Drink: Production and Distribution (Hockwold, 1995), p. 361. 3 F. W. Grube, ‘Cereal Foods of the Anglo-Saxons’, Philological Quarterly, 13 (1934), 140-158, at p. 151. 4 S. J. Crawford, ed., The Old English Version of the Heptateuch (London, 1922), p. 291. Old English ? eru and Modern English Tharf-Cakes ____________________________________________________________________ 119 tion of ibi ‘there’.1 However, a footnote reveals that pær pa is an amendment of the manuscript pæra,2 and I suggest this indicates a form simi lar to peru. The major dictionaries of Old English show how peorf can appear, not only as perf, as shown above, but also as pærf, so the variant vowel is no problem. My suggestion is that the edited phrase etaÍ per pa, ‘eat there then’3 need not have been amended from the manuscript reading etaÍ pæra which I would translate as ‘eat unleavened loaves’. I take pæra be a neuter genitive plural, and to be understood as a partitive genitive.4 I also take it to be a loose translation of a part of the Vulgate text which has been omitted by Crawford, presumably because he believed it to have been Ignored by the translator. After ibi comedite eas, the Vulgate has panes quoque consecrationis edite … ‘eat also the loaves of the consecration …’ If my suggestion is correct, the Old English version was slightly more succinct than the Latin original. The Latin ‘… coquite carnes ante fores tabernaculi et ibi comedite eas panes quoque consecrationis edite …’, meaning ‘… cook the meats before the doors of the tabernacle and eat them there, eat also the loaves of the consecration …’ was translated into Old English as ‘… SeoÍaÍ eowwerne mete beforan Íæs temples dura [and] etaÍ pæra …’, meaning ‘… cook your meat before the doors of the temple and eat the unleavened loaves …’ It is interesting to note that, if I am right, the standard Old English spelling of peorf at Leviticus 8: 2, appears in what I presume to be the spoken form pær just a few verses later at Leviticus 8: 31. C. P. Biggam University of Glasgow * † * 1 Crawford, The Old English Heptateuch, p. 29. 2 The form is pæra in both the manuscripts containing Leviticus: London, British Library, Cot ton Claudius B iv, and Oxford, Bodleian Library, Laud Misc. 509. 3 The Latin phrase ibi comedite eas can be translated ‘eat them there’ because the text has ‘meats’ in the plural, but the Old English has ‘meat’ in the singular. 4 Lass mentions the use, albeit inconsistent, of the Germanic partitive genitive with verbs of eating and drinking. See Roger Lass, Old English: a Historical Linguistic Companion (Cambridge, 1994), p. 237, note 27.