moore - loss of final n in inflectional syllables of middle english

7/27/2019 Moore - Loss of Final n in Inflectional Syllables of Middle English

1/29

Linguistic Society of America

Loss of Final n in Inflectional Syllables of Middle EnglishAuthor(s): Samuel MooreReviewed work(s):Source: Language, Vol. 3, No. 4 (Dec., 1927), pp. 232-259Published by: Linguistic Society of AmericaStable URL: http://www.jstor.org/stable/409257 .

Accessed: 29/12/2011 05:19

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

Linguistic Society of America is collaborating with JSTOR to digitize, preserve and extend access toLanguage.

http://www.jstor.org
http://www.jstor.org/action/showPublisher?publisherCode=lsahttp://www.jstor.org/stable/409257?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/409257?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=lsa


2/29

LOSSOF FINALn IN INFLECTIONALSYLLABLESOFMIDDLEENGLISHSAMUELMOORE

UNIVERSITY OF MICHIGANIn no respect perhapsdo the Southernand MidlandMiddleEnglishtexts of the thirteenthandfourteenthcenturies howgreaterdifferencesthan in the loss or retentionof the final n of unstressednflectional yl-lables. Disregardingall but the gross differences,we can recognisethreetypes of distribution o one or the other of whichthe varioustextstend, in the roughest way and in varyingdegrees,to conform. Thesethreetypesof distributionmaybeillustratedby the OwlandtheNightin-gale,the Cotton Nero A 14 text of the AncrenRiwle,and the LondonEnglishof Chaucer.In the OwlandtheNightingalewe findan approximationo the com-plete loss of final n in all the unstressed nflectionalsyllablesthat de-veloped from the Old English endings -an, -um, -on, and -en: i.e. inboth the singularandpluralof weaknouns,in the weakadjective nflec-tion, in the dative singularandpluralof the strong adjective nflection,in the dative plural of strong nouns, and in the presentsubjunctiveplural,the preterit ndicativeandsubjunctiveplural,the infinitive,andthe past participleof strongverbs. The loss of finaln is approximatelycompletein the singularof weak nouns and in the strong and weak

adjective inflection. There are a few examples, however, of -en asthe pluralendingof nouns that were weak in Old Englishand a verysmallnumberof analogical en pluralsof nounsthat werenot weak inOldEnglish,such as children; ne or two pluralforms also may be in-terpretedas the MiddleEnglishdevelopmentof the OldEnglishstrongdative pluralin -um, e.g. of heoresunnen,858. In verbs the final nis lost in at least 80 per cent of all the plurisyllabicorms that wereen-titled historically o the ending-en.In the Cotton Nero A 14text of the AncrenRiwlethereis an approxi-mately complete oss of finaln in the singularof weak nounsandin thestrong and weak adjective inflection. But final n is nearly alwaysretainedin the pluralform of weak nounsand there are numerousex-

232


3/29

LOSS OF FINAL n IN MIDDLE ENGLISH 233amples of analogical -en plurals of nouns that were not weak in OldEnglish. There are also some -en plurals that may be interpreted asthe Middle English development of the Old English strong dative pluralin -um, but the difficulty of discriminating between dative and accusativein this text makes impossible any more definite statement as to therelative frequency of -en and -e in the dative plural of strong nouns.In verb forms final n is nearly always retained.In the London English of Chaucer there is complete loss of final nin the singular of weak nouns. There is also complete loss of the nasalin both the strong and weak adjective inflection, except for thesurvivalof the Old English ending -an or -um in -self compounds such as myselven, etc.' In the great majority of weak nouns the analogical pluralending -es has displaced the historical -en, but in the very small numberof weak nouns that retain the historical inflection the ending is always-en and never -e; there are a few nouns that have analogical -en plurals.With regard to verb forms it would be rash in the light of our presentknowledge to make a much more definite statement than that neitherloss nor retention of final n was complete in any form of the verb.2Two conclusions, I believe, can be drawn (tentatively, at least) fromthe facts summarised in the preceding paragraphs. From the fact thatsome loss of n occurred in all grammatical forms3and that in texts likethe Owl and theNightingale there was an approximation to complete lossof n we may infer that we have to do here, in part at least, with theresults of sound-change. But from the fact that in all texts, includingtexts like the Owl and theNightingale, the loss of n was more complete insome grammatical forms than in others we may infer that the distributionof forms with and without n was not the result of sound change alone.It seems reasonable to expect that sound change alone would resultin actual speech either in: (1) complete loss of n; or in (2) a distribu-tion of forms with and without n that would correspond to phonetic

1The same exception must be made for the Owland theNightingale, the AncrenRiwle, and all the other earlier Middle English texts in which loss of final n inthe strong and weak adjective inflection is almost or quite complete.2I can only partly agree with Wild's opinion "dass die n-losen formen beiChaucer eigentlich. die normalformen sind, die n-formen aber zur beseitigungdes hiatus von dem dichter fakultativ verwendet wurden" (Die sprachlicheneigentiimlichkeitenderwichtigerenChaucer-handschriftennd die spracheChaucers,p. 296), but it seems fairly clear that the Ellesmere scribe's partiality for the -enforms should not be accepted as representing Chaucer's own usage.3Ormhas no loss of final n in verb forms except the present subjunctive plural,but his spellings cannot be relied on as proving more than that n was usuallyretained in verb forms.


4/29

234 SAMUEL MOOREcategories, not grammatical categories; in written documents, however,the phonetic distribution occurring in speech might very probably be soimperfectly represented that the texts would show (3) a distributionof forms with and without n that would correspond neither to pho-netic nor grammatical categories. The distribution that actually occursis most reasonably accounted for on the hypothesis that some othernon-phonetic process operated along with or subsequent to sound-change.This factor we may call in general terms analogy.

When, however, we inquire as to what analogy operated, it is evidentthat the data we have been considering do not supply the answer. Wemust know when analogy began to affect the distribution of formswith and without n before we can form an opinion as to what these analog-ical processes actually were. Analogy must have begun to operateand must have established a definite trend of linguistic development ata period earlier than that reflected in the literary texts of the thirteenthcentury, for even the earliest thirteenth century texts reflect a latestage of development with respect to loss of n. We must appeal, there-fore, to texts that reflect the speech habits of the twelfth rather thanthe thirteenth century.The texts that are available are twelfth century transcriptions of worksthat were originally composed in the tenth or eleventh century, containedin such MSS as Cotton Vespasian D 14, Bodley 343, Lambeth 487,Laud 636, Hatton 38, and Harleian 6258.4 Such a text is not idealmaterial for linguistic investigation, for its written form cannot bedepended on as reflecting consistently the speech habits of the twelfthcentury scribe who copied it, but is probably determined in some degreeby the written form of the earlier text it was copied from. Nevertheless,these texts, when subjected to adequate linguistic analysis, are capableof yielding information as to twelfth century speech habits that wecannot obtain from other sources.The frequent failure of these twelfth century scribes to reproducethe final nasals that occurred in the Old English texts they transcribedis evidence (especially when taken in connection with what we knowof the later history of final n in Middle English) that final n was notalways pronounced in the spoken language of the scribes. We may

4The Worcester Fragments of the Address of the Soul to the Body are nottranscriptions so far as we know; the Winteney Rule, though a transcription, isin a thirteenth century MS; the "Peterborough additions" in the Laud MS arenot necessarily transcriptions.


5/29

LOSS OF FINAL n IN MIDDLE ENGLISH 235assume that the scribes did not pronounce the final n's they failed towrite, but we are not justified in assuming that they pronounced all thefinal n's they did write. Since these transcriptions were copies of olderMSS in which final nasals were generally retained, it is safer to assumethat a scribe sometimes copied forms with final nasals as they werewritten in his original and sometimes wrote the forms that he pronouncedhimself. Now if loss of final n had occurred indiscriminately and to anequal extent in all the grammatical forms of his speech, the proportion offorms with and without n in his transcription would tend to be approxi-mately the same in the different grammatical categories and the varia-tion in proportions would be the result of his happening to copy the finaln relatively oftener in some forms than in others. The distribution offorms with and without n among the different grammatical categorieswould be a chance distribution, but the differences would not be verygreat. And as the distribution was the result of chance we should expectto find when we compared a number of these texts with respect to therelative frequency of n-forms and n-less forms in the different grammati-cal categories that the distribution would not be the same in all butwould vary from text to text. Moreover, if we consolidated our statis-tics so as to show the proportion of n-forms and n-less forms in thedifferent grammatical categories for all our texts, we should expect tofind that the proportions in the consolidated tabulation would be muchmore nearly equal than those that we found in the individual texts.Now this is both what we do and what we do not find.Table I, printed below, shows the distribution of forms with and with-out n in the following texts:5

1. PeterboroughChronicle, ed. Plummer, Two of the Saxon ChroniclesPar-allel, A.D. 1048-1121, I, 171-250.MS Laud 636; 1122; (538)62. Gospelof Saint Matthew,ed. Skeat.MS Hatton 38; 1154-1189; (903)

6 The date of the MS or the century to which it is assigned is given after theMS notation. This date is usually that given by the editor of the text and is ofcourse only approximate. The number in parenthesis at the end is the totalnumber of forms, with and without n, of all five grammatical categories in eachof the texts. The total for all the texts together is 6760.6This MS is in one hand up to the year 1121and thereafter in various hands.I examined the text from the year 1023 but found no examples of loss of n until1048. My tabulation of the data does not include the "Peterboroughadditions",for which see No. 13 below.


6/29

236 SAMUEL MOORE3. Medicina De Quadrupedibus,ed. Delcourt, Anglistische Forschungen,XL.MS Harleian 6258; c. 1150; (188)14. Peri Didaxeon, ed. L6weneck, Erlanger Beitrige, XII.MS Harleian 6258; XII; (450)75. Early English Homilies, ed. Warner, E.E.T.S., pp. 1-33.MS Cotton Vespasian D 14; XII; (680)6. Passio Beatae Margaretae,ed. Assmann, Bibliothek der angelsdchsischenProsa, III, 170-180.MS Corpus Christi College Cambridge303;XII; (244)7. Winteney Version of Rule of St. Benedict, ed. Schroer.MS Cotton Claudius D 3; XIII, first quarter; (1347)8. Nativitas Sancte Marie, ed. Assmann, Bibliothek der angelsdchsischenProsa, III, 117-137.

MS Bodley 343; XII; (171)19. Twelfth Century Homilies I, ed. Belfour, E.E.T.S., pp. 2-76.MS Bodley 343; XII; (794)810. In Die Pentecosten;De Octo Viciis; ed. Morris, Early English Homilies,I, 87-119.MS Lambeth 487; XII; (286)11. De Initio Creature,ed. Morris, Early English Homilies, I, 217-231.MS Cotton Vespasian A 22; c. 1200; (124)912. History of the Holy Rood Tree, ed. Napier, E.E.T.S.MS Bodley 343; XII; (190)1013. PeterboroughChronicle,ed. Plummer, op. cit., "Peterborough additions,"I, 29-33, 35-37, 39, 52f., 65, 71, 115-117, 127, 144, 163, 183, 198f., 202,203, 205-207, 209, 234, 238, 241, 245f., 247.11MS Laud 636; c. 1122; (96)14. Fragmentsof Address of Soul to Body, ed. Buchholz, Erlanger Beitrdge,VI, 1-10.MS Worcester; XII; (123)15. Twelfth CenturyHomilies II, ed. Belfour, E.E.T.S., pp. 78-140.12MS Bodley 343; XII; (626)

TTexts 3 and 4 appear, according tothe information kindly furnished me

by J. P. Gilson, Esq., of the MSS Department of the British Museum, to be inthe same hand.s Texts 8,9, 12, and 15 are all, according to information kindly furnished by Mr.George Watson of Oxford, in the same hand.9J. P. Gilson, Esq., of the British Museum wrote in reply to my inquiry as tothe date of these homilies that he would be disposed to put them "about thebeginning of the thirteenth century"."' See note 8 above.11As to these see Plummer, op. cit., II, xlv ff., esp. liv. These additions maybe readily recognized from their content and their linguistic characteristics.12This part of the MS is, as stated above in note 8, in the same hand astexts 8, 9, and 12,but the distribution of forms with and without n in this part isnotably different from that in the other three texts.


7/29

LOSS OF FINAL n IN MIDDLE ENGLISH 237Table I is divided into three parts, A, B, and C. In A is given thepercentage of loss of final n for each of the fifteen texts in five grammati-

cal categories. The grammatical categories used are the singular ofweak nouns, the weak adjective, the dative singular and dative pluralof the strong adjective, the plural of weak nouns, and the dative pluralof strong nouns. The percentage was obtained by dividing the totalnumber of forms of a given grammatical category into the number. offorms of that category that had no final n. E.g., text 1 has 122 examplesof the singular of weak nouns in which the n is retained and 7 in whichthe n is lost; the percentage was obtained by dividing 129 (122 plus 7)into 7, giving .054, which was counted as .05. The bottom line of A,marked "Total", shows the percentage of loss of final n in each of thetexts for all five grammatical categories taken together. The last columnof A, marked "Total", shows the percentage of loss of final n for eachof the five grammatical categories in all the texts taken together. Thepercentages were obtained as before, by dividing the total number offorms with and without n into the number of forms without n.In B the data on which A is based are presented so as to show thepercentage of loss of n in the singular of weak nouns as in A and in thestrong and weak adjective taken together and in the plural of weak andstrong nouns taken together, the percentages being obtained by thesame method as was used for A.In C the same data are presented but further assembled so as to showthe percentage of loss of final n in the singular of weak nouns, the weakadjective inflection, and the strong adjective inflection on the one handand in the plural of weak and strong nouns on the other, the percentagesbeing obtained by the same method as was used before.The loss of n in verb forms in these texts is not shown in the table.The majority of texts show no loss of n in verb forms (or a single examplethat may be due to miswriting). The Hatton Gospels and Medicinade Quadrupedibus how a very small percentage of loss of n in verbs andPeri Didazeon, the Winteney Rule, De Initio Creature, and the Peter-borough Additions show a considerable loss of n. It is evident that thefrequency of loss of final n in verbs has no consistent relation to the fre-quency of loss of final n in the noun and adjective inflection.I have stated above that the distribution of forms with and withoutn in the twelfth century texts both is and is not of the kind that we shouldexpect to result from chance. When we examine the percentage of lossof n in the first three grammatical categories (singular of weak nouns,weak adjective, and strong adjective) we see that the percentages are


8/29

238 SAMUEL MOORETABLE I11

Texts1 23 4 1112131 1 Total9si i , 1 0 1 1

AWk.noun, s....... 05 .20 .21 .25 .61 .49147 .28 .37 .38 .63 .84 .68 .84 .72 .38Wk. adj.......... .36 .26 .14 .14 .50 .51 .28 .48 .50 .53 .67 .63 .63 .83 .86 .50St. adj. d. s. pl...... 20 .18 .36.49 1.2419.29 .69.64 .79.59.72.921.00.88 .43Wk. noun, pl

.........01 .06 .14 .21 .07 .04 .26 .00 .07 .00 .00 .00 .25 .08 :28 .11St. noun, d. pl...... .04 .13 .18 .14 .02 .14 .24 .14 .14 .02 .14 .05 .71 .07 .36 .15

Total............ .13 .17 .22 .30 35 .40.41 .54 .65.702B

Wk. nouns, s....... 05 .20 .21.25 .61 .49 .47 .28 .37 .38 .63 .84 .68 .84 .72Wk. and t. adj......29 .21 .30 .33 .37 .33 .29 .59 .5 .63 .63 .651.72 .92 .87Nouns, pl......... 03 .11 .16 .17 .03 .11 .24 .09 .1301 .11 .04 .42 .07 .34C

Adj., wk. nouns,.. .20 .21 .25 .27 .421.39 .35 .47 .52 .57 .63 .70 .70 .89 .85Nouns,p...........3 .1116.17031124091301110442 07341sThe data on which Table I is based were obtained from a single count of theforms with and without n in the texts. The percentages should therefore beregarded merely as approximate. I made a second count, however, of the formswith and without n in two of the texts (6 and 10) and found that the percentagesobtained from the second count were for both texts the same in three of thegrammatical categories as those I had obtained from the first count, but that forboth texts the percentages obtained for the strong and weak adjective in the

second count were a little smaller (between .01 and .04) than those obtained inthe first count. As to the method used in collecting the data: (1) doubtfulcases were usually decided in favor of retention of n in the singular of weak nounsand the adjective inflection and in favor of loss of n in the plural of nouns; (2)loss and retention of inflectional n was counted in stressed syllables (e.g. OEtweo, twgon) as well as unstressed syllables; (3) in the great majority of thetexts I counted as weak nouns only those that were historically weak in OldEnglish and disregardedanalogical -en plurals of nouns that were not historicallyweak, but in two or three texts (1, 11, and possibly 6) I counted the analogical-en plurals among the plurals of weak nouns; the effect of including the analogical-en plurals was to make the percentage of loss of n in the plural of weak nounssomewhat smaller for these texts than it should have been, but as the percentageswere very low anyhow (.01, .00, and .04) the effect produced was small.


9/29

LOSS OF FINAL n IN MIDDLE ENGLISH 239not consistently greater or consistently smaller for one grammaticalcategory than for another. If we rearrange them so that the greatestpercentage will always stand first and the smallest last we find that thethree grammatical categories occur in all the six arrangements that itis possible for them to appear in and that the number of times each ofthe six arrangements occurs is: 1, 3, 4, 1, 2, and 4. The percentagesare very far from being equal, either for the same grammatical categoryin the different texts or for the three grammatical categories in the indi-vidual texts, but when we examine the total percentage of loss of n forthe three grammatical categories in all the texts taken together we findthat the consolidated percentages are very much more nearly equal thanthe percentages for the individual texts. The percentages for the firstthree grammatical categories give no indication that loss of n was morefrequent in one grammatical category than in another; they are thosethat might result from a chance distribution.When we examine the percentage of loss of n in the last two grammati-cal categories (plural of weak nouns and dative plural of strong nouns)we find that in four texts there is a greater percentage of loss in the weaknouns and that in eleven texts there is a greater percentage of loss in thestrong nouns, but it would be rash to conclude from this numericalpreponderance of eleven to four that loss of n was actually greater inthe dative plural of strong nouns than in the plural of weak nouns.Moreover when we examine the total percentage of loss of n for the twogrammatical categories in all the texts taken together we find againthat the consolidated percentages are very much more nearly equal thanthe percentages for the individual texts. The relative frequency ofloss of n in these two grammatical categories considered only in relationto each other appears to be the result of chance rather than of any otherfactor.When we consider the percentages for the first three grammaticalcategories in comparison with those for the last two grammatical cate-gories, however, we see that the relations are altogether different fromthose that we found in comparing together the percentages of the twogroups taken by themselves. The percentage of loss of n is consistentlygreater in the first three grammatical categories than in the last two.This is most strikingly shown in the totals showing the percentage ofloss of n for the five grammatical categories in all the texts taken together,which are .38, .50, and .43 for the first group and .11 and .15 for thesecond. The really important and significant differences, however, arefound in the percentages for the individual texts. We see from part C


10/29

240 SAMUEL MOOREof Table I, showing the total percentage of loss of n for the first threecategories taken together and for the last two taken together, that inevery text the percentage of loss is substantially greater in the firstgroup than in the second.In part B of Table I, showing the percentage of loss for each of thetexts in the first, in the second and third, and in the fourth and fifthgrammatical categories, the quantitative relations are similar to thosethat appear in part C. In every text but one the percentage of loss forthe last two grammatical categories is substantially less and in mosttexts very much less than for either the first category or the second andthird taken together. In text 7 (the Winteney Rule), however, the dif-ference is less than the others we have found, the percentages for thistext being .47, .29, and .24 respectively.The difference of distribution of forms with and without n in the twogroups is most conclusively shown in part A of Table I. In three textsonly, 3, 4, and 13, is the greatest percentage in the second group largerthan the smallest percentage in the first group. In text 1 the greatestpercentage in the second group, .04, is not very much less than the small-est percentage in the first group, .05. In text 7 neither percentage inthe second group is very much less than either of the two smallestper-centages in the first group. In the other ten texts the greatest percent-age in the second group is substantially greater than the smallest per-centage in the first group.The fact that in the twelfth century MSS the loss of final n is verydecidedly and on the whole consistently greater in the first three gram-matical categories than in the last two cannot reasonably, I believe, beregarded as the result of chance. It is with much more probabilityaccounted for as a reflection of the speech habits of the scribes whocopied the MSS. But we must not lose sight of the fact that in theseMSS the distribution of forms with and without n is controlled by threeindependently variable factors: the written form of the text that wascopied, the speech habits of the scribes who made the copies, and chance.The rather small differences of distribution that we find between Mede-cina de Quadrupedibus(3) and Peri Didaxeon (4), both in the same handof Harleian 6258, may very well be the result of chance. So may thedifferences that we find between Nativitas Sancte Marie (8) and TwelfthCentury Homilies I (9), both in the same hand of Bodley 343. But thedifferences between these two texts and 12 (History of the Holy Rood-tree), also in the same hand, seem rather to reflect differences of distribu-tion in the texts that were copied. And Twelfth Century Homilies II


11/29

LOSS OF FINAL n IN MIDDLE ENGLISH 241(15) has a distribution decidedly different from that of 12 and stillmore different from that of 8 and 9 that seems again to reflect a differ-ence of distribution in the text the scribe was copying in this part of theMS. The effect of chance upon the distribution is probably eliminatedto a considerable degree in parts B and C of Table I, but the relativeinfluence of the two other factors, the speech habits of the scribes andthe written form of the texts copied, cannot be accurately ascertained.14It seems clear however from the data presented in Table I that in thetwelfth century as in the thirteenth century the percentage of loss offinal n was decidedly greater in the singular of weak nouns and in theadjective inflection than in the plural of weak nouns and the dativeplural of strong nouns. Our data will not take us much farther thanthat, for tho they seem entirely reliable as a rough measure of the rela-tive frequency of the loss of n in the different grammatical categoriesthey cannot be depended on to furnish an absolute measure of theextent to which loss of final n occurred in the twelfth century. But itseems safe to infer that final n was lost to a greater extent than is re-vealed by the percentages of Table I and that those percentages magnifysomewhat the difference that actually existed between twelfth and thir-teenth century Middle English. In the twelfth century as in the thir-teenth the unequal distribution of loss of n cannot be accounted for asthe result of sound-change alone but points to the cooperation of somenon-phonetic, analogical process. But we are still a considerable wayfrom having reached the beginning of loss of n in Middle English andtherefore can form no opinion as to what this analogical process was.For further light on the problem we must go to texts of the eleventhcentury.The twelfth century texts that we have been examining make, as theyare arranged in Table I, a continuous series that begins with the Peter-boroughChroniclein which the total percentage of loss of n in the fivegrammatical categories is .13 and ends with Twelfth Century Homilies

14 A particularly clear case is the Lambeth MS, which shows a very muchsmaller loss of n in the two homilies based on known Old English originals thanappears in any other part of the MS. Here it seems very probable that thedifference in frequency of loss of n is due in part to linguistic differencesin thetexts copied. It is possible of course that in this MS and in others we have usedthe scribe was not influenced at all by his own speech habits, but reproducedhisoriginal letter for letter. But the conversion of texts in which final n was alwayswritten into texts in which final n was largely omitted must reflect at some pointor at various points in the process of transmission the changing speech habitsof the scribes who copied the texts.


12/29

242 SAMUELMOOREII in which the total percentage of loss is .72.1 I cannot supplementthis series with a corresponding series of eleventh century texts showinga loss of n between .00 and .13. In fact I have found no eleventh centurytext that shows a loss of n great enough to be expressed in terms of per-centage at all. There are, however, a certain number of texts thatcontain unmistakable evidence of some loss of n and that probablybring us about as close to the beginning of the process as it is possiblefor us to come. The data obtained from these texts are presented inTable II below. This table shows the numberof examples in the twenty-four texts of loss of n in the five grammatical categories considered beforeand in the various forms of the verb (including past participle of strongverbs).,e (For the examples themselves and the texts used see theAppendix to this paper.) The last column, marked 'Total', shows thewhole number of examples of loss of n in each text, the first numberbeing the total for the first five grammatical categories only, the secondincluding verbs also.Perhaps the most striking feature of this exhibit is that some loss offinal nasals occurred in all the grammatical forms, including verbs (inwhich, as we have seen, the final n was especially stable thruout boththe twelfth and thirteenth centuries). This fact confirms the hypoth-esis that, however much the development was affected subsequentlyby analogical processes, the loss of final n was in the first instancethe result of sound-change. But the distribution of forms with andwithout n in the eleventh century texts has no relation to the phoneticenvironment of the n17 and we are therefore led to inquire whether itis the kind of distribution that might result from chance or whether itis an unequal distribution that corresponds to grammatical categories.Table II shows the absolute oss of nasals in the grammatical categories

1i No chronological sequence of the MSS themselves is of course implied in thearrangement of the texts; the arrangement was adopted for clearness and con-venience of presentation.16Not all of these MSS are of the eleventh century; some are early twelfthcentury MSS. As to the method used in collecting the data on which Table IIis based: (1) only clear cases of loss of n are included, except in the plural ofnouns, where a probability was accepted; (2) no texts are included except thosethat have at least two cases of loss of n in adjective or noun; (3) a few apparentexamples of loss of n in the dative singular of strong adjectives were disregardedbecause of the possibility of their being instrumental instead of dative singular;(4) loss of n in particles and uninflected forms was not regarded; (5) the loss ofn was (I think) disregarded in plural verb forms immediately followed by a pro-noun subject.17The following data indicate roughly the phonetic environment of the loss


13/29

LOSS OF FINAL n IN MIDDLE ENGLISH 243TABLE II

Weak Wk. St Wk. St.Texts nouns, adj. adj. nouns, nouns, Vb. Totala. pl. d. pl.

1. Alfred, Sol.............. 2 1 1 7 4 (11)2. Elf., Horn. Jud........... 2 1 2 (3)3. Alf., L. Sts.............. 1 1 1 2 (3)4. AElf.,O. T., J............ 3 4 2 95. Chron., MS D........... 5 3 1 8 (9)6. Chron., MS F........... 3 3 1 1 87. Bede, Ecc. Hist.......... 1 1 4 2 (6)8. Boethius................ 2 1 39. Byrht., H................ 1 6 1 810. Defensor, L. S........... 1 1 211. De Nat. S. M............ 1 1 212. Exameron................ 1 2 1 413. Gospel of Nic............ 4 5 1 1 10 11 (21)14. Har. of Hell............. 1 2 1 3 (4)15. Heptateuch..... .... 4 2 616. Hornm.n John 13......... 2 217. Ikceboc. ................ 2 218. Malchus................. 2 1 319. Rect. Pers ............. 1 3 4

20. Rule of Chrod........... 1 2 2 3 (5)21. Rule of S. B ............. 1 1 222. Sign Lang............... 1 1 223. St. Guthlac ............. 3 2 524. Wulfstan................. 2 1 1 3 (4)Total ........................ .37 46 6 5 6 28 100 (128)Percentage of total (nounsand adj. only)............ .37 .46 .06 .05 .06

of n in the 105examples that occur in 17 of the texts used for Table II (the textsnot covered being 3, 9, 10, 17, 19, 21, 22):Final n lost before consonants (not 'weak h') 60" " " " vowels 15" " " " 'weak h' 4it i" ' " clear pause 17c" " " " comma 9

105The relative frequency of loss of n before consonants and vowels in these 105cases is about what we should expect from the fact that in the average passage ofOE the number of words beginning with a consonant is about four times as greatas the number of words beginning with a vowel.


14/29

244 SAMUELMOOREconsidered. There are 100 examples of loss of n in the first five gram-matical categories included in the table; this total is the sum of thenumbers 37, 46, 6, 5, and 6. To compare these numbers with each othercan give us no information as to whether there are relativelymore ex-amples of loss of n in some grammatical categories than in others.For twelfth century texts we have a reliable measure of the relativefrequency of the five grammatical categories in the following data show-ing the total number of forms (both with and without n) in the texts onwhich Table I is based:

Wk. noun sg. Wk.adj. St. adj d. s. pl. Wk. noun pl. St. noun d. pl. Total1286 1538 1239 501 1290 585418.22 .26 .21 .09 .22

A similar, but not necessarily identical, distribution of the five gram-matical categories would be expected in the eleventh century texts; for,tho the linguistic differences between the two sets of texts are not of akind that would materially change the proportions, differences in thecontent of the material might very well affect the distribution. More-over we cannot expect to find in a total of 100 forms the exact propor-tions that occur in a total of 6000. In order to obtain an approximatemeasure of the relative frequency of the five grammatical categories ina total of 100 forms taken at random from the eleventh century texts,I took at random five passages from each of twenty texts used forTable II and counted in each passage the first five examples of any ofthe five categories.'9 The distribution in these five samples of 100forms each was as follows:

18 These are the data for all the twelfth century texts except 1, 6, and 11;that is I have included all the texts in which I did not count as weak plurals thenouns with analogical -en plurals (see note 13above). For all the texts (including1, 6, and 11) the corresponding percentages are almost identical, namely:.22 .25 .21 .09 .22

For the four longest texts (2, 5, 7, 9) they are about:.20 .25 .23 .08 .24

1"The first passage was the beginning of each text, the other four were takenat random. I used twenty texts rather than twenty-four because twenty isdivisible into 100,and the twenty I used were those I had immediately at hand;they included all but 9, 12, 19, and 22. The first sample is less strictly a randomsample than the others; it is not surprising therefore that it departs more widelythan any other from the average distribution of the five samples. It is to be


15/29

LOSS OF FINAL n IN MIDDLE ENGLISH 245

Wk. noun sg. Wk. adj. St. adj. d. s. pl. Wk. noun pl. St. noun pl.

11 30 30 4 2513 25 24 7 3114 31 27 9 1917 28 24 9 2213 30 23 7 2768 144 128 36 124.14 .29 .26 .07 .25

All of these distributions are strikingly different from the distribu-tion of the categories in 'the total of 100 forms showing loss of n:

37 1 46 6 5 [ 6This wide departure from the norm cannot reasonably be attributed tochance and we are therefore led to inquire what other factor or factorsmay have affected the distribution.We may first note that the eleventh century texts show a relativelylarge loss of the final nasal in the singular of weak nouns and the weakadjective inflection as compared with the plural of weak nouns. A con-venient measure of the inequality of distribution is the relation betweenthe aggregate loss of final n in the first two grammatical categories takentogether and the loss of n in the third category. The following tabula-tion shows this relation in comparison with the distribution of these threegrammatical categories that we find in our five samples and in the twelfthcentury MSS:

Wk. noun Wk. adj. Wk. noun Ratioag. pl.

Sample 1.......................... 11 30 4 41:4 10.25Sample 2......................... 13 25 7 38:7 5.42Sample 3......................... 14 31 9 45:9 5.00Sample 4......................... 17 28 9 45:9 5.00Sample 5......................... 13 30 7 43:7 6 14Total....................... 68 144 36 212:36 5.8912 Cent........................... 22 26 9 48:9 5.33Loss of n ........................ 37 46 5 83:5 16.60

understood that the data given for the relative frequency of the five gram-matical categories in both eleventh and twelfth century texts has to do only withthe forms that in OE ended in -an or -um.


16/29

246 SAMUELMOOREThe ratio between the number of examples of loss of n in the first twogrammatical categories and the third is very much greater than the cor-responding ratios in our five samples and the twelfth century MSS.This fact seems to me to indicate a probability that in the speech of(say) the last quarter of the eleventh century the distribution of formswith and without n in these three grammatical categories was not theapproximately equal distribution that should result from the operationof a mere sound-change but one which had already been modified byanalogical processes. The statistics on which this inference is basedseem to me too limited in amount to be conclusive; they need to besupplemented and checked by additional data from other eleventhcentury texts that I have overlooked and from the texts of which Ican-vassed only a part. When we consider the evidence, however, in thelight of the conclusive evidence of unequal distribution of loss of n inthese three grammatical categories that we found in the twelfth centurytexts, the hypothesis that analogy had already begun to affect the dis-tribution before the end of the eleventh century seems entitled to pro-visional acceptance.We find a wide departure from the normal distribution also in thedative singular and dative plural of the strong adjective and the dativeplural of strong nouns, the loss of the final nasal being much less fre-quent in these forms than we should expect from our other samples.The fact that these forms ended in -um in Old English might suggestthe explanation that the late Old English or very early Middle Englishchange of final m to n lagged behind the loss of n, so that the loss ofthe nasal was less frequent in the -um forms than in what we may callthe -an forms.20 Now the evidence of a large number of tenth andeleventh century MSS shows unmistakably that the beginning of thechange of final m to n definitely preceded the loss of final n;21and the

20 I am disregarding for convenience the fact that the OE dative plural endingof the weak adjective and the weak noun was -um and not -an. To separatecompletely the -um and -an forms in these two grammatical categories wouldresult in impracticable complexities and I trust that no serious error is involvedin merging them.21The presentation of this evidence I must reserve for a later paper (to beentitled Earliest Morphological Changes in Middle English) which will dealwith the relative chronology of the change of m to n, the loss of n, and othermorphological changes in very early Middle English. 1 must reserve also forthis later paper the discussion of Jordan's opinion (HandbuchdermittelenglischenGrammatik128,cf. also 152): 'Der ae. ersatz der endung des dat. pl. -urndurch-an ist nicht lautgesetzlich, sondern analogisch'.


17/29


18/29

248 SAMUELMOOREOn the basis of this hypothesis we may now inquire as to what analogi-cal process or processes could have brought about the unequal distribu-

tion of forms with and without n that we actually find in the early MiddleEnglish texts.The analogical processes that operate at a particular period are neces-sarily those that result from the particular morphological pattern thatcharacterises the speech habits of that period. Now, for our presentpurpose at least, the morphological pattern of Middle English between1075 and 1125 can be defined with considerable exactness. Althothe levelling of the unaccented vowels a, e, o, and u to a vowel that after1200 was usually written e had already been carried out, analogicalchanges, which ultimately effected a complete transformation of theinflectional pattern of nouns, adjectives, and pronouns, had barely begunto operate.24 In fact the only systematic analogical change affectingadjectives or nouns that was carried out before the end of this periodwas the development of analogical final e in the nominative singular offeminine 5- and ji-stems that ended in a consonant in Old English.With this exception the inflectional forms of nouns in the earliest MiddleEnglish were the Old English forms modified only by the sound-changesthat took place at the end of the Old English or beginning of the MiddleEnglish period.25Now one of these sound-changes was the loss of final n in unstressedsyllables. As to the conditions under which this sound-change occurredwe have no satisfactory evidence. The simplest explanation of thelater development would be to assume that the loss of final n was origi-nally a combinative sound-change, dependent on the phonetic environ-ment of the nasal. At least it seems certain that the loss of n must havebeen conditioned in someway, for if entirely unconditioned it would haveresulted in the complete loss of n not only in nouns and adjectives but inverbs as well. Analogical processes alone could not have preserved orrestored inflectional n in the face of a sound-change that was tendingunrestrictedly to its suppression. The result of the operation of aconditioned sound-change respecting loss of n in unstressed syllables,however, would be that each word would have two forms, with andwithout n. Assuming that the sound-change was phonetically con-ditioned, the original distribution of the forms with and without n would

24The evidence for this statement and for the statement contained in thesentence following will be presented in the paper referred to in note 21 above.

25The only important qualifications that need to be made to this statement willappear in what follows.


19/29

LOSS OF FINAL n IN MIDDLE ENGLISH 249have depended on the phonetic environment of the word. Under cer-tain conditions, however, this primary distribution would be capable ofsuch modification by analogical processes as would result eventuallyin the establishment of an entirely different distribution that was inde-pendent of phonetic environment.It would seem as if analogical processes could modify the primary dis-tribution of forms with and without n in either of two ways. Loss ofn would be retardedin a particular grammatical category if the n-formwas so much more consistent than the n-less form with the patternof the inflectional system as a whole that speakers would tend uncon-sciously to use the n-form even in phonetic situations where n tended tobe lost. Loss of n would be acceleratedin a particular grammaticalcategory if the n-less form was so much more consistent than the n-formwith the pattern of the inflectional system as a whole that speakers wouldtend unconsciously to use the n-less form even in phonetic situationswhere n tended to be retained. But if neither the n-form nor the n-lessform was more consistent than the other with the inflectional patternthe use of the form with or without n would be left to the determinationof phonetic tendencies alone. Our specific inquiry therefore is this:in what grammatical categories was either the n-form or the n-less formmore consistent than the other with the pattern of the Middle Englishinflectional system between 1075 and 1125?The eleven principal types of noun inflection in Old English and theearliest Middle English are shown in the tabulation printed below. Theendings alone are given, the symbol 0 (zero) being used to indicate thelack of inflectional ending. Forms that in Old English ended in -anand -um are given in the Middle English column as having either -enor -e. The eleven types are numbered arbitrarily for convenience ofreference and are identified by examples of nouns conforming to each ofthe types. By far the great majority of nouns are inflected accordingto one or other of these types and I believe that the tabulation, thoit does not include all the Old English declensions, is entirely adequatefor exhibiting the analogical potentialities of the inflectional system.2"Even a casual examination of this synopsis shows the unsystematiccharacter of the Old English noun inflections. The only endings thatare positively indicative of a particular case and number are -es, -as,-ena, and -um. Apart from a definite grammatical context a noun ofunknown gender ending in -e might be any case of the singular, a nounending in -u might be nominative singular, accusative singular, nomina-

21See note 29 below.


20/29

TABLE IIIOld English Earliest Middle English

Sing. Plur. bing. Plur.1. wulf,mas. Nom. -0 -as wulf -0 -esGen. -es -a -es -eDat. -e -um -e -en, -eAcc. -0 -as -0 -es2. ende, mas. Nor. -e -as ende -e -esGen. -es -a -es -eDat. -e -urn -e -en, -eAcc. -e -as -e -es3. lim, neut. Nom. -0 -u lim -0 -eGen. -es -a -es -eDat. -e -um -e -en, -eAcc. -o -u -o -e4. word,neut. Nom. -0 -0 word -0 -0cynn, neut. Gen. -es -a cynn -es -eDat. -e -urn -e -en, -eAcc. -0 -O -0 -05. rice, neut. Nom. -e -u rice -e -eGen. -es -a -es -eDat. -e -urn -e -en, -eAcc. -e -u -e -e6. lufu, fem. Nom. -u -a lufe -e -eGen. -e -a, -ena -e -e, -eneDat. -e -um -e -en, -eAce. -e -a -e -e7. hwil, fem. Nom. -0 -a hwile -e -esynn, fem. Gen. -e -a synne -e -eDat. -e -um -e -en, -eAcc. -e -a -e -e8. sunu, mas. Nom. -u -a sune -e -eduru, fem. Gen. -a -a -e -eDat. -a -um -e -en, -eAcc. -u -a -e -e9. hunta, mas. Nom. -a -an hunte -e -en, -e

Gen. -an -ena -en, -e -eneDat. -an -um -en, -e -en, -eAcc. -an -an -en, -e -en, -e250


21/29

LOSS OF FINAL n IN MIDDLE ENGLISH 251TABLE III-Concluded

Old English Earliest Middle EnglishSing. Plur. Sing. Plur.

10. tunge, fem. Nom. -e -an tunge -e -en, -eGen. -an -ena -en, -e -eneDat. -an -um -en, -e -en, -eAcc. -an -an -en, -e -en, -e11. Zare, neut. Nom. -e -an ere -e -en, -eGen. -an -ena -en, -e -ene

Dat. -an -um -en, -e -en, -eAcc. -e -an -e -en, -e

tive plural or accusative plural, and a noun ending in -a might be nomina-tive singular, genitive singular, dative singular, nominative plural, geni-tive plural, or accusative plural. Even if the gender and declension ofthe noun were known, its ending was seldom indicative of a particularcase and number. The nominative and accusative plural were identicalin all declensions and the nominative and accusative singular in most ofthem. The dative singular was identical with the accusative singularin most declensions and in six of them the genitive singular and dativesingular were alike. And the only case endings that were definitelycorrelated with gender distinctions were nominative singular -a, nomina-tive-accusative plural -as, and nominative-accusative plural -u; theonly case ending that was positively gender-distinctive independent ofa grammatical contex was -as.One grammatical concept, however, was very consistently expressedby the Old English noun inflections, namely plurality. In all the caseforms of all the declensions, with very few exceptions, the plural form isdifferent from the corresponding singular form. The only exceptions(marked in the tabulation by bold-face type) are the uninflected nomina-tive and accusative plural of word and cynn (4), the genitive plural ofsunu and duru (8), and the accusative plural of hunta and tunge (9 and10). In eleventh century texts, however, nouns like word frequentlyhave a nominative-accusative plural ending in -u, formed on the analogyof nominative-accusative plural forms like limu (3).27 These analogicalplurals in -u are particularly frequent in Defensor's Liber Scintillarum,2827Sievers 238, note 1.28 Ed. Rhodes, E.E.T.S. For a list of these plurals in u see p. xv of the in-troduction.


22/29

252 SAMUEL MOOREbut they occur also in a number of other texts. We may note furtherthat sunu and duru are the only nouns of their inflectional type (8)that are used with any frequency (or at all) in the plural. The accusa-tive plural forms of types 9 and 10 were therefore the only plural formsthat were consistently deficient in the expression of plurality in thelanguage of (say) about 1050.29Unsystematic tho they were, the Old English noun inflections hadpersisted for several centuries with relatively little change and wouldprobably have remained stable for an indefinite period so long as themorphological pattern was not radically modified by sound-change.But the sound-changes that occurred at the end of the Old Englishperiod shifted the traditional patterns to such a degree that they becameexposed to the action of analogical processes to which the Old Englishpatterns were immune. Which brings us back to the question raisedabove: in what grammatical categories was either the n-form or the n-less form more consistent than the other with the pattern of MiddleEnglish noun inflection between 1075 and 1125?The genitive singular of the Middle English weak nouns hunte, tunge,and ere ended in -en or -e, the nominative singular in -e. The n-lessform of the genitive singular conformed to the pattern of three otherinflectional types, 6, 7, and 8, the conformity being expressed in thefollowing proportion :0

hunte (gen. s.) : hunte (nom. s.) :: lufe (gen. s.) : lufe (nom. s.)hwile " " : hwile " "sune " : sune

The dative and accusative singular of Middle English hunte and tungeand the dative singular of ereended in -en or -e, the nominative singular29A detailed examination of the other OE types of noun inflection (abouttwelve in number) would show that this statement applies with unimportantqualifications to them also. Such an examination is superfluous for the readerwho is familiar with the minor details of OE noun inflection and would be merelytedious for the reader who is not. I shall therefore say only that (1) in someof the minor inflectional types (e.g. the masculine wa-stems and the root con-sonant-stems) the plural was consistently distinctive; (2) in others (e.g. theneuter wa-stems, masculine i-stems, and r-stems) an earlier nominative andaccusative plural that was identical with the singular was regularly or frequentlydisplaced in later OE by distinctive analogical plural forms; (3) the others

(e.g. in-stems like bra'du and ipa-stems like lengpu) included only abstractnouns for which the grammatical category of number was a matter of indif-ference.30 In the proportions given here and later the analogical formation is indicatedby italics.


23/29

LOSS OF FINAL n IN MIDDLE ENGLISH 253in -e; the n-less form, therefore, conformed to the pattern of five otherinflectional types, 2, 5, 6, 7, and 8. This conformity may be expressedin the following proportion:hunte (dat., acc. s.) : hunte (nom. s.) :: ende (dat., acc. s.) : ende (nom. s.)wite " " " : wite " "lufe " " " : lufe " "hwile " " " : hwile " "

sune " " " * sune "

The support given to the n-less forms by these other inflectional patternswas so strong that it seems reasonable to conclude that the effect of ana-logical processes would be to accelerate the loss of n in the singular ofweak nouns."3

The analogical processes that were actually effective in favoring thesurvival of either the n-form or the n-less form in the plural of nounswere not of a kind that can be so easily expressed in the form of a pro-portion. When we examine the earliest Middle English inflectionaltypes as they are given above in our tabulation, we observe that pluralityis expressed with much less consistency than by the Old English inflec-tional patterns. In no less than five types of inflection (4, 5, 6, 7, 8)the nominative and accusative singular and plural were identical. Fromthis fact we might infer that the analogyhunte (nom., acc. pl.) :hunte (nom., acc.s.): :rice (nom.,acc.pl.) :rice (nom.,acc.s.)lufe " " " :lufe " " "

hwile " " " :hwile " " "sune " " " :sune " " "

would favor the survival of the n-less form rather than the n-form andthat the loss of n would therefore be accelerated in the plural of weaknouns. The evidence of the eleventh and twelfth century texts, how-ever, shows unmistakably that in spite of the effects of sound-change inshifting and blurring the traditional patterns of inflection, plurality innouns was still a living grammatical concept that tended to receiveformal expression in speech. We have already noted the fact that in

31 The analogical support given to the n-less forms of the dative and accusa-tive singular was stronger than that given to the n-less form of the genitivesingular. Moreover, in five of the early ME inflectional types (1, 2, 3, 4, 5) thegenitive singular (ending in -es) was different from all the other cases of thesingular. On a priori grounds we might therefore expect the -en forms to besomewhat more stable in the genitive singular than in the dative and accusativesingular.


24/29

254 SAMUELMOOREeleventh century MSS analogical plurals in -u tended to replace theuninflected plural forms of nouns like word (type 4). Still more signifi-cant is the fact that analogical n-plurals begin to occur in the eleventhcentury MSS and occur with increasing frequency in those of the twelfthcentury. These analogical n-plurals are most frequent in types 3, 7,and 8, but occur also in types 4, 5, and 6. In other words, the speech-feeling for the expression of plurality was so strong that the n-forms ofthe nominative and accusative plural of weak nouns not only survivedthe tendency to loss of n, but themselves furnished the basis for analogi-cal developments that may be expressed in the proportionsynnen (n., acc. pl.) : synne (n., acc. s.) :: hunten (n., acc. pl.) : hunte (n., acc. s.)sunen " " " : sune " " "

If we accept the conclusion that a feeling for the formal expressionof plurality retarded the loss of n in the nominative and accusative plu-ral of weak nouns, it necessarily follows that the same explanation ac-counts for the retention of the n-form of the dative plural of*weaknouns.The same analogical tendency operated to retard the loss of n in thedative plural of strong nouns. For in all the strong inflectional types ofearliest Middle English the dative singular ended in -e and the dativeplural in -en or -e; loss of n in the dative plural, therefore, would havedestroyed all distinction between singular and plural in that case.The adjective inflection in earliest Middle English furnished little orno basis for analogical processes that could either accelerate or retardloss of final n in the weak adjective and the dative singular and pluralof the strong adjective. The inflectional pattern of nouns furnished abasis for analogical processes that favored loss or retention of final npartly because of diversity of inflectional types and partly because finaln in the plural of nouns had a functional value that favored its survival.The two types of adjectives, those ending in -e and those ending in aconsonant, were nearly identical, however, in their inflection and theslight differences between the two types were not such as furnished abasis for analogical developments. Moreover, final n had very littlefunctional value in the adjective inflection either in Old English or inthe earliest Middle English. In extremely few forms of the weak adjec-tive in Old English was the final n distinctive of case or gender and onlyin the nominative plural and neuter accusative plural was it distinctiveof number. In fact the form of the weak adjective was seldom distinct-ive of anything except the fact that the adjective was not strong. Allforms of the weak adjective except the nominative and accusative singu-


25/29

LOSS OF FINAL n IN MIDDLE ENGLISH 255lar neuter of adjectives ending in -e and the dative plural were differentfrom the correspondingform of the strong adjective. In earliest MiddleEnglish (with loss of final n) the strong and weak form of the adjectivebecame identical also in the nominative singular masculine and femineof adjectives ending in -e and in the dative singular masculine and neuter,accusative singular feminine, and nominative and accusative plural ofall adjectives. The distinction between the strong and weak form ofthe adjective was still made in earliest Middle English in most forms ofadjectives ending in a consonant and in some forms of adjectives endingin -e. I am unable to find in the inflectional pattern of earliest MiddleEnglish the basis for any analogical processes that could have preservedthe distinction more completely. We may therefore infer that loss offinal n was neither retarded nor accelerated by analogical processes inthe weak adjective and that loss of n in this grammatical category wasthe result of sound change alone.3The dative singular masculine and neuter and the dative plural of thestrong adjective had much more distinctiveness of function in Old Eng-lish than the weak adjective forms. Both were distinctive of case andthe dative singular masculine and neuter form was at least negativelydistinctive of gender, tho neither form was distinctive of number. Asthe result of change of final m to n, the levelling of unstressed vowels,and the loss of final n the Old English ending -um was reduced to -ein earliest Middle English. The n-less form of the dative plural thusbecame identical with the nominative and accusative plural form. Then-less form of the dative singular, however, was still distinctive ofgender and case to a considerable (tho somewhat reduced) extent, es-pecially in adjectives ending in a consonant. Distinctiveness of func-tion would have been completely preserved if loss of n had been arrestedby analogical processes, but the inflectional pattern of the strong adjec-tive in earliest Middle English seems to have furnished no basis for suchanalogical developments. I infer therefore that loss of final n in thedative singular masculine and neuter and dative plural of the strongadjective was the result of sound-change alone.The problem of loss of final n in verbs lies outside the scope of the pres-

2' An examination of the tabulation given above will show that loss of nwas relatively more frequent in the singular of weak nouns than in the weakadjective. This fact would seem to confirm the hypothesis that loss of n wasaccelerated in the singular of weak nouns, but was left to the determinationof phonetic tendencies in the weak adjective. The fact, however, needs theconfirmation of more statistical data before we are justified in regarding it assignificant.


26/29

256 SAMUEL MOOREent paper, but cannot be entirely ignored, for the fact that final n was somuch more stable in verb forms than in noun and adjective forms thru-out the eleventh and twelfth centuries is obviously relevant to the con-clusions apparently indicated by the data presented here in regard toloss of n in adjectives and nouns. These conclusions are: (1) that lossof final n in late Old English and earliest Middle English was initiallya combinative sound-change that resulted in double forms, with andwithout n, distributed according to the phonetic environment of the n;(2) that this primary distribution was then modified by analogical proc-esses that accelerated loss of n in the singular of weak nouns and re-tarded loss of n in the plural of nouns; (3) that loss of n in the adjectivewas neither retarded nor accelerated but was the result of sound-changealone. If these conclusions are correct it would seem to follow that lossof final n in verb forms was either retarded by analogical processes or elsethat the phonetic conditions under which loss of n occurred were con-ditions to which verb forms were less frequently subject than noun andadjective forms. The analogical potentialities of the verbal patternseem less than those of the inflectional pattern of nouns but they are notwholly lacking.33 As to whether verb forms were less frequently sub-ject than noun and adjective forms to the phonetic conditions underwhich loss of n occurred we can form no opinion until those phoneticconditions have been ascertained. I would suggest as a mere possibil-ity that if final n should have been lost before consonants (or beforecertain consonants) and preserved in hiatus and beforea pause the verbforms might be found to occur with such frequency before a pause as tomaterially retard the loss of n. But if we could account for the stabilityof final n in verb forms thruout the eleventh and twelfth centurieswe should then have to answer the further question why the n-formsdid not remain equally stable thruout the thirteenth and fourteenthcenturies. It is clear that the highly important problem of loss of nin verbs must be solved by other methods and other materials thanthose used in the present paper.

13 The chief analogical potentialities of the verbal pattern tending to retardloss of n seem to be the inflected forms of the infinitive and past participle. Themonosyllabic verbs (e.g. OE dan, bion, soon, gan), which tho few in number werevery frequent in use, were also a conservative factor. So also was the factthat in all types of weak inflection the plural form was different from the cor-responding singular form, and that in all finite verb forms final n had the samefunction, the indication of plurality.


27/29

LOSS OF FINAL n IN MIDDLE ENGLISH 257APPENDIX

The data upon which Table II is based are as follows:31. Alfred, Soliloquies, ed. Hargrove, pp. 1-40; Cotton Vitellius A 15; XI orearly XII.

P]eawa11.8; hwilce 17.19; eaga 17.21; sunne 20.19; ]ince 10.15; forlete13.5; lufia 19.10; begyte 23.7; ofercumme 34.18; nmbbe36.9; habbe 36.92. AElfric,Homily on Judith, ed. Assmann, Bibliothekder angelsachsischenprosa,III, 102-116;Corpus Christi College Cambridge 303; XII.soya 15; lyfigenda 346; nolde 813. AElfric,Lives of Saints, ed. Skeat, E.E.T.S., I, 10ff., lines 1-428, II, 168ff.,lines 1-473; Cotton Julius E 7; XI.hlafdige I, 28.47; fyrmeste II, 208.304;Penode II, 206.2634. Xlfric, On the Oldand New Testament;Judges, ed. Crawford, E.E.T.S., pp.15-51, 64-75, 401-417;Laud 509; XI.sida 172; heretoga 394; witega 666; deriendlica 813; ungewmpnode 1127;lichama 1132;fagera 1144;sceamlica 1248;witerwinna Judges, VIII: 285. Chronicle,MS D, ed. Thorpe, pp. 248-350;Cotton Tiberius B 4; XI and XII(several hands).gewuna p. 256;wilda p. 264;gerefa p. 266; waestrenap. 270;Hamptuniscap. 292; utlaga p. 314;maga p. 328; papa p. 328; onfange p. 2746. Chronicle,MS F, ed. Thorpe, pp. 249-329; Cotton Domitian A 8; XI or XII.masse p. 267; gehadode p. 267; yldesta p. 275; feawa p. 301; Su6sexa p.303; husbunda p. 313; utlaga p. 3197. Bede, Ecclesiastical History, ed. Schipper, Bibliothek der angelsdchsischenprosa, IV, 1-11, 18-50; Cambridge University Library Kk 3, 18; XI.utagangende 414; arlease 469; agife 127; afyrhte 1082;bregde 1085; gefre-mede 10878. Boethius, ed. Sedgefield; continuous parts of text from MS B, pp. 7-11, 14,21, 26f., 33f., 46f., 48ff., 51f., 57f., 64, 67f., 69, 71, 73f., 79ff., 89, 94ff., 101, 105,115f., 124, 125f., 135f., 141, 146f.; Bodley 180;XII (early).willa 11.19; sunne 141.19; micle 146.99. Byrhtferth's Handboc, ed. Kluge, Anglia VIII, 298-301, 312-314; 322-324,335-337; Ashmole 328; XI.

mwela301.3; lengtentima 312.22; getyddusta 313.3; gewuna 313.34; wyn-sume 313.37; fulfremede 322.24; getydde 335.3; iudeisce 335.4710. Defensor's Liber Scintillarum, ed. Rhodes, E.E.T.S., pp. 1-43, 213-222;Royal 7 C. iv; XI.tima p. 9; leofesta p. 13.3"The approximate date of the MS or the century to which it is assigned isgiven after the MS notation. The date given is usually that assigned by theeditor or editors, but I owe to the courtesy of Mr. George Watson of Oxfordmyinformation as to the date of MSS 9 and 11 and to the kindness of J. P. Gilson,

Esq., of the British Museum my information as to the date of MS 18. In theexhibit of forms showing loss of n the verbs are placed at the end of each group.The references are usually to page and line or to line alone. Accents and marksof quantity are not reproduced.


28/29

258 SAMUEL MOORE11. De Nativitate Sanctae Mariae, ed. Assmann, Bibliothek der angelsdchsischenprosa, III, 117-337;Hatton 114, formerly Junius 22; XI.

sceape 94; Peowa 20112. Exameron, ed. Crawford, Bibliothek der angelsdchsischenprosa, X; CorpusChristi College Cambridge 302; XI.35eorce 136; ungelarede 152, feorta 200; susle 49913. Gospel of Nicodemus, ed. Hulme, P.M.L.A.A., XIII, 473-515; Cotton Vitel-lius A 15; XI.eorpa 473.6; hebreisca 473.20; wytega 481.13; goda 483.5; leofe 487.17;onhangena 487.28; nama 487.32; lychama 493.21; gebygede 495.20; wizier-winna 501.15; sylfa 513.16; cute 475.4; andswarode 475.26; 477.5; 477.13;speca 481.25;wolde 483.7; nyste 485.26;bydda 489.6; gefagenogde (sic) 491.8;eode 495.19

14. Harrowingof Hell, ed. Hulme, ModernPhilology, I, 610-614; Corpus ChristiCollege Cambridge 41; XI.halga p. 610;ordfrumap. 610; stranga p. 611; cyme p. 61115. Heptateuch,ed. Crawford, E.E.T.S., Exodus, pp. 212-285; Cotton ClaudiusB 4; XI.nama II:10; laesse XVIII:22; leohtra XVIII:26; nama XX:7 (twice);oxa XXI:3616. Homily on John XIII:1-30, ed. Assmann, Bibliothek der angelsdchsischenprosa, III, 151-163;Bodley 340; XI (?).Pweorre 50; lichamlice 5917. Lvceboc, ed. Leonhardi, Bibliothek der angelsachsischen prosa, VI, 84-87;Harleian 55; XI.aswollena 85.2; untruma 85.718. Malchus, ed. Assmann, Bibliothek der angelsdchsischenprosa, III, 199-207;Cotton Otho C 1; XI.nacode 304; lichama 304; efenpeowa 38219. Rectitudines Singularum Personarum,ed. Liebermann, GesetzederAngelsach-sen, I, 444-453; Corpus Christi College Cambridge 383; XII.lande p. 446; Candelmaessep. 446; lande p. 447; landa p. 44820. Rule of Chrodegang,ed. Napier, E.E.T.S., pp. 1-63; Corpus Christi CollegeCambridge 191; XI.late 8.21; anwearda 29.3; twy 33.13; naebbe6.20; belimpe 19.2021. Rule of St. Benedict, ed. Schrier, Bibliothek der angelsdchsischenprosa, II,9-23, 32-36, 45-64; Corpus Christi College Cambridge 178; XI.willa 22.2; slapule 47.1722. Sign Language (Indicia Monasterialia), ed. Kluge, Internationale Zeitschriftfar allgemeineSprachwissenschaft,II, 118-129;Cotton Tiberius A 3, f. 97;XI.swypra 123.25; samlocone 128.423. St. Guthlac,ed. Gonser, Anglistische Forschungen,XXVII; Cotton VespasianD 21; XI.ece 106.34;unmeattra120.64; ichama 146.7;Hadde 154,heading; abodysse158, heading

35Crawford prints the text of Hatton 115 and gives all the variants of theother MSS, including CCCC 302.


29/29

LOSS OF FINAL n IN MIDDLE ENGLISH 25924. Wulfstan's Homilies, ed. Napier, homilies 36, 44, 51-53, pp. 172-175, 215-226, 274-277; Cotton Tiberius A 3; XI.feawe 221.33;godcunde 275.16;godcunde 276.9; abera 223.11

[The observation that the different numbers of one and the same case are regu-larly distinct in form seems to me of interest. With few exceptions it will holdfor Sanskrit-but jas, vifvapas (s. pl.) also before voiced consonants senm(s)(pl.) send (s.) and Vedic quci (d. pl.), Cuci, puru, karma, karmd (s. pl.)-forGreek, for Latin-but moles, sedes, dies, res-and for Lithuanian to judge fromLeskien, Lit. Lesebuch 153. Apparently it was a well marked tendency in IEinflection, though I cannot recall a statement to that effect, and do not find itwhere it might be expected: Brugmann, Grundr.22. 2. 114, overstresses theexceptions, while Hirt, Idg. Gram.3.38, makes no allusion to the matter. Theserviceableness of such distinctions is obvious; and also that, because of theaccompanying verb, they are least so in the nominative-the case which furnishesall of the above exceptions.Jespersen, Lang. 429-31, cites from languages of 'barbarous races' a numberof instances in which a people is able to name this or that species of tree, orparrot, but is without a word for 'tree', or 'parrot'. Indo-European may fairlybe added to the list: our linguistic ancestors could (and must) say 'accusative-singular', 'instrumental-plural', etc. etc., but for 'singular' and 'plural' theyhad no formal expression in the inflection of the noun.-G. M. B.]

moore - loss of final n in inflectional syllables of middle english

Documents