38840847 English Morphology

8/2/2019 38840847 English Morphology

1/12

\

Chapter 2EnglishMorphology

' , . ,

MORPHEMES

[

t us begin by assuming that we are archeologists who wish to makeinferences about the three primitive drawings and strings ofgeometric shapes contained in figure 2.1. Examine each onecarefully; the three iIJustrations are not intended to form asequential narrative.! Let us hypothesize that the geometric

" 0 C lFigure 2.1: Some primitive drawings accompanied by geometric symbols.

shapes are symbols which comment on the drawings and that we canprovide rough meanings fo r the three sequences of symbols as follows:

2.18 A 0 Cl 'man bites dog'b 0 I:l. t= l 'dog bites man'e 6 * =::J 'man bites camel'

Lnguis[s call the preliminary statements of meaning glosses and placethem in single quotation marks. Each word is said to be a gloss for a

'This use of geometric shapes i$ an adaptation of a problem discussed by Cook (1967,pp. 21-2 2).


2/12

ENGLISH MORPHOLOGY

given geometric shape; the string of words c ~ n likewise be called a glossfor the string of shapes. The data in 2.1 constitute a corpus, or a text.Examine the corpus carefully. Which geometric symbol occurs on lyonce in the corpus? Which gloss occurs qn ly once? The star must therefore mean 'camel'. Which geometric symbol occurs in but twO of thethree strings? Which gloss occurs on ly in the two corresponding linesof glosses? The circle must therefore mean 'dog'. The triangle andrectangle occur in all three strings but the position of the triangle is different in 2. tb; the position of 'ma n' likewise changes. The triangle musttherefore mean 'ma n' and, by eliminatio n, the rectangle must mean'bites', We have thus assigned a gloss to each symboL

But what in fac t does mean 'dog' or 'man' in the communication sysrem we have been examining? 1 have said that a circle means 'dog'. But ina tec hnical mathematical sense there really are no circles in figure 2. 1.Even the naked eye can deteCt that the "circles" are imperfectly drawnand that the tWO figures perceived to be circles are different in size andshape. Must we therefore retract our inference that a circle symbolizes'dog'? Must we,say that there are tWO, and maybe even more, symbolsfor 'dog': a small circle- like figu re with an indentation on the upper rightand a slightly larger circle-like figure with a protrusion o n the upperright? What of different colored circles? Circles drawn with chalk ?Circles traced in sand? Is each a different symbol for 'dog'? It seems nOt,A cirdt symbolizes 'dog', This means that anything perceived to be acircle can be a symbol for 'dog' in the communication system we areconsidering. Thus [he meaningful unit that sy mbolizes 'dog' is not anyparticular circle but the abstraction "circle." Let us move on to anotherarray of data that will shed more light on the abstract nature of linguisticsymbols.Let us consider a corpus from a hypothetical human language. Thedata in 2.2 look more like the kind of data that linguists usuaUy wor kt h ~ o n l y the specialist needs to decode archeological inscriptions,

as we had to in the previous ex ample. We can assume that the data andthe glosses have been elicited from someo ne who knows the languageand who also knows enough English to provide us with glosses. Linguists call such a person an informant. Examine carefully the corpusbelow.

2.28 rekpasi 'birds sang'b darpasi 'children sang'c dus! '3 chi ld sang'd lobasi 'robins sang' losi 'a robin sang 'f rofsi 'a boy sang9 disi 'a gir l sang'

23MORPHEMES24

In 2.1 we were al ready provided with discrete symbols, each of whichwe initially presumed to correspond to one of the words in the gloss.In 2.2 we have an uninterrupted sequence oflerrers which we must somehow segment: we must discover which letters or combinations of letterscorrespond to which words or parts of words in the gloss. The procedureis nonetheless quite similar to that emp loyed in the analysis of 2.1. Theessence of the procedure is to note where cbanges in the gloss corre.spond to changes in the text, or inversely, where recurrent sameness inthe gloss corresponds to sameness in the text. Examine the corpus again,this time drawing lines between groups of letters that seem to correspond to a word or part of a word in the gloss, Only 'sang', you probablynoticed, occurs in every gloss, and only si occurs in every line of the text.We may initial ly look for a segment in the text which corresponds exactly to the indefinite article 'a', but we will find no recurrent correspond.ing segment in the text. However if we take 'a girl' as a meaningful unitwhich occurs only once in the glosses, we will notice that di likewiseoccurs only once and in the corresponding line in the text. Similarly,rof seems to correspond to 'a boy'. We are left with the following portions of the Corpus unsolved:

2.3a tekpa 'birds'b datpa 'children'c cl " 'a child'd loha 'robins'e 10 'a robin'Note that dat occurs in 2,3b and 2.3c, and 10 in both 2.3d and 2.3e, Bur2.3b and 2.3d contain an additional syllable. Is there a unit of mean ingexpressed in the glosses of 2.3b and 2.3d that might account for the additional syllable in the corresponding lines of the text? Yes, Both nounsare plural. We may tentatively conclude, then, that pa in 2.3b and ba in2.3d mean something like 'more than one' and that dat is the singular'child' and 10 the singular 'rob in '. We can now further infer chat the pain 2.3a likewise means 'plural', leaving lek to mean 'bird'. Let us now listthe meaningful segments in the corpus along with the individual glossof each:

2,48


3/12

ENGLISH MORPHOLOGY

The most important fact to nOte from ou r analysis is that in 2.4f we havetwo segments for on e meaning. Linguists have found this is not unusualin language. Indeed, even in the English glosses, we have a similar simacion: the notion 'plural' is represented by the J in 'birds' and 'robins' butnot so in 'children'. We can often find some system in the alternationof segments which have the same meaning. When the segments are pronounced quite similarly as in 2.4f, the alternation is usually conditionedby the pronunciation context. A hypothesis that emerges from the dataof 2.2 is that the segmentpa occurs after consonant sounds, and the segment ba after vowel sounds. Thus we may consider pa and ba to be systematically determined variants of on e and the same unit of language,which-like the "circle"-is an abstraction. Just as the symbol for 'dog'in 2.1 was concluded to be anything perceived to be a circle, so in 2.2 theunit that means 'plural' is anything perceived to be . , , perceived to bewhat ? Mathematics provided us with a readymade perceptual abstraction for use in 2,1, but apparently no ready-made abstraction exists forou r present purposes. Before discussing further how linguists treatsuch facts and resolve this need, let us examine yet another array of data,this time in English:2.58 boys, bags, rails, trains, hamsb tacks, bats, cuffsc bushes, watches, judgesd oxen, brethren, childrene teeth, geese, feet

f data, criteria, alumniEach English word in 2.5 c o n t a i ~ s tWO meaningful segments. Since thedata are in English, 1 have no t provided glosses. Such glosse:s cou ld bein another language or they might even be pictures-of a boy, a train,and so forth. Th e gloss of the second segmenr in each word is of course'plural'. Th e spelling of this segmenr is the same in 2.5a and 2.5b: an J .But in 2.5c to 2.5f each set of words manifests a different spelling ofplural'. Furthermore, jf we attend closely to the pronunciation of thewords, we will note that although 2.5a and 2.5b both en d in the letcer i,the pronunciation of this letcer is nOt the same in 2.5a as it is in 2.5bj in2.5a the letter s representS a "z" sound, whereas in 2.5b it representS an"s" sound, Thus all of the following mean 'p lural': (1) the "z" sound of2.5a, (2) the "s" sound of 2.5b, (3) the tJ of2.5c, (4) the en of2.5d, (5) thevowels spelled .ee as opposed to 00 in 2.5e, and (6) the various vowelendings in 25f. This is why the linguist glossing an English text wouldwrite 'plural' in the gloss of every word in 2 . 5 ~ And anyone knowingEnglish will think 'more-than-oneof-whatever-it-is'when he hears any of

25WORDS26

the words in 2.5, Linguistics must take account of this recurrent sameness of gloss or meaning. It does so, in this case, by establishing an abstract unit of language that encompasses all the various sounds that mean'plural'. JUSt as the notion "circle" encompasses an almost limidess varietyof entities perceived to be circles, so a label like ti , for instance, can beestablished to encompass any and all sounds or combinations of soundsthat can be glossed 'plural' or that evoke the idea of plurality in the mindof someone knowing English. In this book, I label this particular abstraction es. I could alternatively have given it a number or simply called it'plural'. Any such abstract label for a group of different sounds havingthe same gloss is what linguists call a morpheme.

WORDSWe have seen that a morpheme is an abstract meaning label attached

to the smalles t meaningful segmen ts of sound in a language; usually on emorpheme label will group together two or more segments with differentpronunciations bOt with the same meaning. Ou r choice of the plural morpheme as an example from English should have made clear that a morpheme is not necessarily a word, Users of English feel that the pluralmorpheme is part of a word. And this is true. In fact, it is convenient tothink of the domain of morphological analysis as the word: segmentingand labeling the morphemes that make up words, and subdividing morphemes into types according to the roles they play in forming words.When we treat the morphology of English, we study how words are puttogether. And this will be our concern throughout the remainder ofthis chapter.

I f morphology treats the composition of words, then [he discussionof morphemes contributes to defining what a word is. Thus this entirechapter constitutes such a definition. But even before beginning a detailed presentation of the facts, we should at least have a tentative definition in mind. Fortunately, the intuitions of native users of English arequite adequate; these are manifested in, and reinforced bv the conventions of English spelling: a word is a unit of language, which: ~ h e n written

primed, is separated from other units by a space. Bu t let us try todiscover and make explicit some sources of these conventions. Tw ofacts should be kept in mind: (1) that every word Contains at leas t on emorpheme, and (2) that many words contain tw O or more morphemes.Th us, once the morphemes of a language are ide ntified and listed amajor task of morphology is to state how the morphemes group i ~ t o words.


4/12

ENGLISH MORPHOLOGY

Consider the two-word senten


5/12

ENGLISH MORPHOLOGY

Consider the word subtract, which is composed of twO morphemes,sub and tract. Even though their meanings are abstract, they are fairlyconsistent; sub meaning 'down' or 'under' and tract 't o pull' or 't o take'.Now, imagine a linguist analyzing an English corpus and looking forot her occurrences of the morpheme sub and the morpheme tract. I f hehas access to a large corpus, he would find a very large number of wordscontaining sub. Here is JUSt a sampling:

2.98 subdue, subject, submerse: submit, subscribe. subsidize, subsist,substitute, subsume, subvertb subjoin, sublease. sublet, submerge, subserve, sub tendHowever, even with access to the same la rge corpus, he would find rela-tively few words containing the morpheme tract. Here is a virtuallycomplete list:

2.10 attract, contract, detract. distract, extract, protract, retractTh e lists in 2.9 and 2.10 poim ou t an important difference be twee n suband tracl. I f we view the word subtract as composed of two posilions(one occupied by sub, the o the r by Iract). we immed iate ly notice thatfor each position ther e are only certain morphemes that can replace theone already there and still make up a genuine word. Very many mor-phemes can occupy the position of tract bu t only relatively few canoccupy [he position of Sllb. This is not an isolated phenomenon true onlyof the word subtract: Consider the word express, composed of the mor-pheme ex and the morpheme press. Ho w many morphemes can youthink of which can substitute fo r press after ex? Here is a partial list:

2.118 exalt, excise, exclude, excuse, exempt, exhume, exonerate, ex-pand, expect, expedite, expel, explain, explode, expOfl, ex-punge, extol, extort. exrract. extricate, exudeb exclaim, exchange, expatriate, expose, extendNow, how many morphemes can you think of that can substitute for exbefore prm? Here is a virtually complet e list:

2.12 compress, depress, oppress, repress, suppressMany other examples like these can be cited, examples which show thatsome morphemes occupy a position in wo rds where there is a relativelyunlimited potential for substituting other morphemes (linguists callsuch morphemes roots) and that other morphemes occupy a positionwhere there is a relatively limited potential for substitution (these arecalled affixes) . In ou r two examples, affixes came before the roots. Affixes may also follow roots. Consider the word traction, composed of the

29SOME ENGLISH PREFIXES30

morpheme tract and the morpheme ion. It would be easy to show thatmany morphemes can replace tract in front of ion (fo r example, tension,fusion, lesion, fJersiofJ, portion, plosion , fission), and thus tract is a rOOt. bu tthat very few morphemes (perhaps only able and or) can replace ion aftertract, and thus ion is an affix. Affixes preceding roots are called prefixes,and affixes following roOtS are called suffixes. Bur prefixes may alsoprecede other prefixes, as in t he word McompreHion where M is a prefixthat precedes another prefix, com, and bo th precede the root press. Andsuffixes may also follow o ther suffixes as in the word tractability, whereily is a suffix that follows another suffix, abil, and both fo llow the roottract. Furthermore, a word may have both prefixes and suffixes (fo r ex-ample, distractible, retraction, expressible, comprmion). linguists generallycite prefixes with a hyphen after them (for example, sub and ex-) andsuffixes with hyphens in froot of them (for example, ion and -ible).Let us consider another phenomenon related to the classification ofmorpheme types. Hopefully, you will have asked yourself why the ex-amples in 2.9 and 2.11 were divided into tWO groups under a and b.Reexamine these lists of examples, trying to determine a basis fo t thedivision. None of the roOts in 2.9a or 2. 11 a can function as a word-none has a free distribution. Th e roots in 2.9b and 2.11b. on the otherhand, can function as words. For these reasons, the roots in 2.9a and 2.11aare called bound roots, and those in 2.9b and 2.llb are called freeroots. Here is a summary of the terms and definitions introduced inth is section:

2.13a Roo/s: morphemes whic h occupy a position in a. word where thegrea.tes t potentiality for subsdtution exists(i) Bound: incapable of functioning as a word(ii) Fru: capable of functioning as a wordb AjftxlJ: morphemes which occupy a position in a word whereonly a limited potentiality for substitution exists

(i) Prefixes: affixes which precede roots and possibly otherprefixes(ii) SujfiXtS: affixes which follow roots and possibly other su f-fixes

SOME ENGLISH PREFIXESExamine che sampling of English prefixes in 2.14. Several others ace

treated in an exercise at the end of this chapter. I group the prefixes intosix categories based on their meaning. These meaning categories are notexhaustive, bu t they do encompass a la rge number of English prefixes.Tw o examples of prefixes from each meaning type are given. As you


6/12

ENGLlGH MORPHOLOGV

examine the example words listed beside each prefix, notice that whena prefix is added to a root, the composite word is generally (he same partof speech as the root without the prefix. Notice also that, while prefixesdo not normally have as many pronunciation variams as did. the -es morphem.e we considered above. there is enough variation to reinforce ou rdefinition of a morpheme as an absttact meaning label which may,grouptOgether a variety of pronunciations. In this regard, note especially thedifferent pronunciations of poly - in polyglot and polygamy and the differen t pronunciations of hyPel'- in hyperactive and hyperbole.

2.14a Number bi-poly-b Time poSt-

pree Place inter-

d Degree hyper.ultra-

e Privation. un,- dis-

f Negation. non-un,-

bifocal, bilingual, biceps, bicycle~ J ; s Y i J ! 6 i Q P O l y W J ~ ypostwar, postelecrion, posrciassical,postpOneprewar, preschool, pre-19rn century,premaritalinternationaJ, intercontinental, intertwine,interact, intermarry, in'terdependence,interplaysubway, subsection, subconscious, sublet,subdivide, subcontract, subliminaJhypercritical, hyperactive, hypersensitive,hyperboleultravio let, ultramodern, ultraconservative,ultraliberal, ultramarineundo, untie, unzip, unpack, unleash,unhorsedisconnect, dis'infect, disown, dispose (of),disheartened, discolored, discontent,displeasure, dissatisfactionnonconformist, nonsmoker, nonpolirical,nondripunfair, unwise, unforgettable, unassuming,unexpected

DERIVATIONAL AND INFLECTIONAL SU FFIXESAlthough I used prefixes as examples to distinguish between rootsand affixes, I did note that affixes may also be suffixes and gave several

examples. In fact, the majority of English affixes are suffixes. Let us nowexamine the behavior of some English suffixes.

Consider the verb root treat in the sentences of 2.15. I t occurs In2.15a without a suffix and in 2.15b and 2.15c with suffixes.

31DERJVA TlONAL AND INR.ECTIONAL SUFFIXES32

2.15a They treat their employees well.b That treatment helped cure the disease.C He treated his students fairly.

Note that the combination of the verb root treat and the suffix -ment in2.15 b results in a word that is a noun. (It is tcue that there is a noun root,treat, in English, as in the sentence He bought me a treat, bU[ if we examine other words in English with the suffix -ment, such as arrangementand amazement, we find that -ment is clearly affixed to a verb root.) Wemay conclude that when the suffix -men! is affixed to a verb, the resultis a noun. No w notice that when the verb rOOt treat combines with thesuffix -ed in 2.15c the result is still a verb. Thus -men l changes the partof speech of a rOOt to which it is affixed, but -ed does not. Here is anotherexample. The noun rOot symbol occurs without a suffix in 2.16a but withsuffixes in 2.16b and 2.16c:

2.16a Purple is a symbol for royalty.b White flags symbolize surrender.C Both the circle and the trian.gl.e can be symboh.

Note that the combination of the noun root symhol with the suffix jzein 2.16b results in a verb. Other examples of a noun root and -ize combining to form a verb are hospitalize and vaporize. We may conclude thatwhen the suffix :ize is affixed to a noun, the result is a verb. But noticethat when the noun root Jymbol combines with the suffix -s in 2.16c theresultant word is still a noun. Thus -ize changes the part of speech of aroot to which it is affixed, but -s does not. Here is a list of the twO rOOtsand four suffixes we have just discussed with summary comments inparentheses:

2.1 78 treat (a verb root)b symbol (a noun root)C -mem (a suffix that chan,ges verbs to nouns)d -ize (a suffix chat changes nouns 0 verbs)e -ed (a verb suffix-that means 'past')f s (a noun suffix-that means 'pluraJ ')

Suffixes like -ment and -he which change the part of speech of words arecalled derivational suffixes. Suffixes like -ed and -s which do not changethe part of speech of words are called inflectional suffixes. Bu t let uslook further into the behavior of these two rypes of affixes.I noted ear lie r in this chapter that a root may have twO or more suffixes atrached to it. Here are tWO examples:

2.18a treatmentsb symbolized


7/12

ENGLISH MORPHOLOGY

No tice that in 2. 1S each word is composed of a root, a derivational suffix,and an inflectional suffix. In 2.1Sa, -ment changes treat to a nou n, andthus the -s that indicates plural in nou ns can be added. In 2. 1Sb, -hechanges symbol to a verb, and thus the -ed that indicates past in ve rbscan be added. Suppose we tried to reverse this orde r of affixation, reasoning as follows: if tnat is a verb rOOt, then -ed can be added givingtreated, which is still a verb. Now if -ment changes verbs-to nouns, whynOt add it to th e verb treated giving the noun treatedment? But the combination of morphemes ind icated by the spelling treatedmml is in factimpossible in English. And so is the combination '"symbolsize, whichmight be formed by adding first an inflec tiona l and then a derivationalsuffix co the rOOt symbol. Such forms are nOt unreasonable, nor would theybe without a ce rtain usefulness. Bu t they are evidenrly nar used in English. And thus we have a second criterion for distinguishing betweenderivatio nal and inflectional suffixes: inflecdonal suffixes can follow derivational suffixes in words, but derivadonal suffixes cannot follow inflecdonal suffixes. Not ice that the in flectional suffixes -ed and -s occuronly last in the words listed in 2 .1 9.

2.19a treatmentb sy mbolizec treatedd sy mbolse sy mbolizationf symboli2able9 treatmentsh symbolizedsymbolizabil itysymbolizationsBefore go ing on to discuss another criter ion of distinction. between

derivational and inflectional suffixes, le t me introduce twO useful termsand illustrate them by reference to the words listed in 2 .19. We havecalled treat and symbol roots. Thus we are able to say that treatment andsymbolize are each composed of a roo t and a derivational suffix and thattreated and symbols are each composed of a root and an inflectional suffix.But if we wish ( 0 discuss the linguistic form to which the last suffix isattached in 2. 1ge to 2. 19 j, we may face a terminological handicap. Consider ;ymbolizable and treatments. We cou ld say of course that -able is asuffix which derives an adjective from a rOOt that is a verb (for example,breakable) and from a roOt-plus-another-affix that is a verb (for example,symbolizabk). Or we could ignore the fact that -able is added just to a roo tin breakable and to a roor-plus-another-affix in symbolizable and say that-able derives adjectives from worth that are verbs whether the words

33DERIVATIONAL AND INFLECTIONAL SUFFIXES34

are simple roars or roots plus one or more ot her suffixes. Neverthelessit seems useful to have terms that can refer ei ther to roOts or to com bi-nations o f roots and affixes, and that can indicate whether a de rivationalOt inflectional suffix can be affixed to such a morpheme or combinationof morphemes. Linguists often use the term stem to refer to any morpheme or combination of morphemes to which a given inflectional suffixcan be affixed. Thus in the word treated, the root treat can also be calledthe stem of the -ed inflect ional suffix. Bu t in the word symbolized, thestem of the -ed inflectional suffix is symbolize, both a rOOt and a derivatio nal suffix. Similarly, the stem of the -s inflection in the word symbolizations is symbolization, a root plus two derivational suffixes. Similarly,linguistS often use the term base to refer to any morpheme or combination of morphemes to which a given derivational suffix can be affixed.Thus while treat is a root in the word treatmenl, we call it a base relativeto the derivational suffix -ment. In the word jymbolizable the base of thederivational suffix -able is symbolize, a root and anOther de rivational suffix.Similar ly the base of the de rivational suffix -ilY in the word symbolizabiNtyis symbolizable, a rOOt plus two derivational suffixes. H er e are summaryde6n itions of the terms ste m and base:

2.20a Stem: a root or -a root plus other affixes to wh ich a given inflectional suffix can be attachedb Base: a root or a root plus other affixes to which a given deri-vational suffix can be auachedWe now continue our examination of the distinctions between derivationa l and inflectional suffixes. We have seen that derivational suffixes change the part of speech of bases towhich they are attached, butinflectional suffixes do nOt change the part of speech of stems to whichthey are attached. Secondly, de rivationa l suffixes cannot follow inflectional suffixes, but inflectional suffixes must follow all derivational suffixes. Now consider the lists of words in 2.21, which demonstrate a third

cri terion of dislinction.2.218 treatment, arrangement, amazement, judgmentaa "'killmem, "buildment, "spendmentb hospitalize, symbolize, vaporizebb "bookize, "chairize. "houseize

We no ted earlier that the derivational suffix -ment changes verbs tonouns and that -ize changes nouns to verbs. The asterisked words in 2.21demonstrate that ment cannot change jUst any verb to a noun, no r can-ize change JUSt any no un to a verb. In fact, jf you try to add to the listsin 2.21a and 2.21b, you will find that bot h ~ m e n l and -ize operate on on ly


8/12

ENGLISH MORPHOLOGY

very small g roups of verbs and nouns respectively. Now compare thelimitations on the operation of these [W O derivational suffixes with theability of bo th inflectional suffixes we have so far examined to attach tovirtually eve ry stem of a part of speech: every noun has a plural. andevery verb has a past. Thus, derivational suffixes may be affixed to onlyrelatively few bases of a give n part of speech. (This may be why Englishhas several hundred derivational suffixes.) Inflectional suffixes may beaffixed to all stems of a given part of speech. (This may be why. as weshall soon see, English has only eight inflectional suffixes.)

Here is a summary of the thre'" differences between derivational andinflectional suffixes:

2.22 Derivational:a Usually change the panof speech of basesb Cannot followinflectional suffixes

C Usually affix to onlysmall groups of basesof a given pan ofspeech

Inflectional:a Never change the pan ofspeech of stems /b Must follow a1l derivationalsuffixesc Affix to virtually any stemof a given parI of speech

SOME ENGLISH DERIVATIONAL SUFFIXESExamine the eighteen derivational suffixes and example words in 2.23.

They are in six groups of three, based on the change they effect in parrsof speech. (Notat ions like N < V in 2.23c should be read "derives nounsfrom verbs.") Other derivational suffixes are treated in an exe rcise atthe end of this chapter. As you examine the words in 2.23, notice tWOthings: First, some derivational suffixes do not meet all three of our criteria of definition; for instance, the suffixes in 2.23a are affixed to nounbases and (he resultant word remains a noun. Nonetheless the other twOcriteria classify these suffixes as derivational. Second, notice that derivational suffixes, like prefixes, do not ordinarily have many pro nunciationvariants, but they provide reinforcement for our definition of a morpheme as an abstract meaning label because they frequently producepronunciation variants of morphemes in their bases. Notice bow the morphemes sane, vain, and chaste change vowel pronunciation when oilyis affixed: sanity, vanity, chaslilY. An d notice that the "k" sound at theend of pllblic, elastic, and fanatic changes to an "s" sound when derivatio nal suffixes are added: publicilY, e1aslicity,fanaticism.

35SOME ENGLISH DERIVATIONAL SUFFIXES36

2_23a N < N hood neighborhood, sisterhood,bachelorhood, boyhood,brotherhood, childhood,maidenhood, knighthood-ship lordship, township, fellowship,championship, friendship,membership, lectureship, kinship-seer gangster, gamescer, trickster,songster, pu nscer , mobster,prankster, speedsterb N < ADJ/N , -ism idealism, impressionism, fanaticism,dualism, realism, imperialism,

c N< V

... romanticism, patriotism-icy sani.y, vanity, rapidity, banality,elasticity, ability, actualicy, agil ity,chastity, curiosity-ness meanness, happiness, cleverness,usefulness, biuemess, brightness,darkness, goodnessal refusal, dismissal, upheaval, den ial:survival, uial, approval, proposal-er worker, writer, driver, employer,swimmer, preacher, rtaveier,teacher, baker ment arrangement. amazement,puzzlement, judgment,astonishment, treatmentd V < ADJ/N en ripen, widen, deafen, sadden,harden. lengthen, deepen,srren,gthen, neaten-ify

-nee AD) < N/V -ful

-ish-able

ADV < AD)/N -Iy-wise

beautify, diversify, codify, amplify,simplify. glorify, nullifr, Frenchifysymbo lize, hospilalize; publicize,popularize, modernize, idealizeuseful, delightful, pitiful, helpful,careful, awful, rightful, sinful,cheerfulfoolish, selfish, snobbish. modish,hellish, sheepish, Swedish,)ewishacceptable, readable, drinkable,livable. commendable.comfortable, changeablehappily, strangely, oddly,athletically, basically, semanti!=aIlyc 1 o c ~ w i s e , lengthwise, weacbecwise,educationwise '-ward earthward. homeward, eastward


9/12

ENGLISH MORPHOLOGY

ENGLISH INFLECTIONAL SUFFIXESTh e e ight inflectional suffixes of English ace listed in 2.24. Th e word

or phrase given first in parentheses serves bOth as a name and a roughdesignation of meaning. Next is the morpheme label, then words whichend in the inflection. Each line represents a variant pronunciation ofthe suffix itself or a variant it req uires in the pronunciation of certainstems . Examine the lists of examples, preferably by reading them aloud.Notice that 2.24a and 2.24b are noun inAections, 2.24c to 2.24f are verbinflections, and 2.24g and 2.24h are ordinarily adjective inflections.Notice also that many words given to exemplify the past tense inflection in 2.24d are spelled. and pronounced the same as words listed withthe past pacticiple inflection in 2.24e. I sha ll comment below on this andother aspects of 2.24.

2.248 (pluraJ)

b (possessive)

C (present tense)

d (past tense)

e . (past participle)

-es boys, bags, tails, trains, hamstacks, batS, culTsbushes, watches, judges" ~ . f t . e s 4 : ~ j : e s , I t ~ ~ ~ ' i t e s oxen, brethren, childrenteeth, geese, feet 'data, criteria, alwnnisheep, fish, series-poss )im's,)oe's, Mom's, Mr. Moore'sPat's, my wife's, Mike's. the judge's, the witch's, the boss's-pes go/goes, buy/buys, se ll/sellsrake/takes, bat/bats, buff/buffspush/pushes, watch/watchesam/is/are', have/has , do/does-pst stayed, tried, behaved, barredwalked, stopped. stuffed, watchedsighted, banded, remindedchose, rose, spoke, stole, frozethrev.:.,-knew, grew. slewtook, stood, forsookbought, taught, foughtcut, ler, hit, betwent, did, had, made-en stayed, tried, behaved, barredwalked, stopped, stuffed, watchedsighted, banded, remindedbought, let, madetaken, eaten, fallen, known

37ENGLISH INR.ECTlONAL SUFFIXES38

f (present parriciple) -ingg (comparative) -erh (superlative) -est

broken, chosen, spoken, stolenwritten, given, driven, risengone, done, been, seencoming, going, buying, sellingbigger, older, quickerbetter, worsebiggest, oldeST, quickesthest, worst

Th e plural inflection, which we have labeled -e.J, is pronounced with"z," "s," "ez," and "en" sound s, or as a vowel change in a stem, or achange of vowel at the end of a Stem-all of which I discussed ear lierwhen illustrating the nature Of morphemes; but the plural inflection can~ s o have what is called a zero variant: the stems of sheep,fish, and !triesin 2.24a do no t change pronunc iation in any way when the -es inflectionis added, as inJohn bought three sheep. Note also that the -es inAection cancause a pronunciation variant of a stem to appear: whereas the wordswijt, knijt, and thief end in an "f" sound, the stems of their plurals endin a "v" sound. ' .

Early in this chapter I defined a morpheme as an absu'act meaninglabel, and I have reemphasized and illustrated this definitiOn throughoutthe chapter. On e reason I have doqe so is because it has important practical implications. Most of the remaining chapters of this book treatEnglish syntax. As you will soon see, syntax is the study of the interactionof morphemes in sentences. Syntax ignores spe ll ing and .pronunciationvariants. Thus it is important fo r you to learn to ignore variations suchas the examples in 2.24 manifest, and to recall that the plural morphemeis always represented as -et no maner how it is pronounced and even ifit is nor pronounced ar all.

Th e possessive inflection, labeJed -poss, has three pronunciations:a "z" as inJim's, an "s" as in Pat's, and an "ez" as in the judge's. But, again,we ignore such variations when discussing the behavior of this inflectionand represent the morphemes in these three examples as jim-poss, palposs, and the judge-poss.

The present tense inflection, labeled -prs, presents special problems.Ma.nY linguists speak of a "th ird person singular agreement inflection,"whICh affixes to a verb in the present tense when the verb is precededby he, she, il, or a noun tha t these pronouns can replace. This accountsfor the "z," '5," and "ez" pronunciations at the end of some examplewords in the first three lines of 2.24c; it would also account for is, does,and has the fourth line: all three .os-like" variants occur after thirdperson singular nouns. Bu t these "s_like" sounds add no more meaningto a verb stem than it communicates when preceded by T, we, y01l, they,or nouns these pronouns can replace. For this reason, we consider tbe


10/12

ENGLISH MORPHOLOGY

"z," "s," or ez pronunciations to be variants of the -prs morphemewhich alternate with a zero variant that occurs after I, we, and so forth.Thus the verb in both We eal chme and He eaU cheese is analyzed morphologicaUy as eal-prs.Th e past tense inflection, which we have labeled -psI, has a conside r-able Dumber of.pronunciation variants, many of which are exemplifiedin 2.24d. Notice tw O general tendencies: (1) a tendency to indicate 'past'by affixing a "d ," "t," or "ed" pronunciation to the verb stem, as in stayed,walked, and sighted respectively, and (2) a tendency to indicate past bychanging the pronunciation of a vowel in the verb stem, as in chose, threw,and took. Some variants, like bought, combine both tendencies. Someverbs, like CUI, caU for a zero variant. Other verbs seem to manifest thefirst tendency but also to call for a variant in the pronunciation of thestem: in wenl, did, had, and made a "d" or "t" p r o n u n c i a ~ i o n of the pasttense inflection apPears, but the pronunciation of the stem in each casevaries somewhat from the ordinary non past pronunciations (go, do, have,make). Th e morphological representations of wenl, did, had, and madeare go-pst, dO-PSI, have-pst, and make-pSI respectively. Th e same kind ofanalysis is m a d ~ of those verbs with past tense variants that change avowel in the stem: chose is analyzed las choose-pst, threw as throw-PSI, andtook as take-pSI.The past participle in flection, which we have labeled -en, has a familiarvariant pronounced "en" or "n" as in taken, broken, written, gone. Noticethat broken, written, gone, a,nd the words that accompany them in 2.24echange the pronunciation of a vowel in their stem when the -en inflectionis affixed. However, even though the most frequently used verbs ofEnglish call for an "en" or "n" variant, most English verbs pronounce-en .exactly like the past te nse inflection. Notice that the first three linesof example words in 2.24e are the same as the first three lines in 2.24d.We may ask what de termines whether a given occurrence of stayed, orwalked, or made is to be analyzed asJlay-pst or slay-en, walk-pSI or walk-en,make-pst or make-en. In a later chapter, we shall examine detailed criteriafor distinguishing these twO inflections. Here,l wish only to demonstratethat the distinction is a real one and to discuss a simple procedure todetermine which of the twO inflections occurs in a given word. Noticethat in He sighted a UFO the "ed" of sighted dearly means 'past', but inUFOs are sighted every week nothing about the sentence means 'past'.Similarly, the "ed" of lived in He lived there last year is dearly 'past', butthe sentence He has lifled Ihere for Iwo years implies that he is still livingthere, and is thus as much a comment on the present as on the past; weare not like ly to say or hear a sentence like He has lived there last year, andthe reason is of course that the "ed" of lived in this sentence is nOt avariant of the past tense inflection but of -en. Here is a procedure for

39EXERC ISES40

deciding whether a given verb stem endi ng in "ed" has -pstor -en affixed.Find a verb like tak.e, break, write, or go (whose pronunciation of -en isclearly distinct from -pst) and substitute it for the word in question. If,for instance, go were substituted for live in He has lived there, it wouldimmediately be dear that He has gone there not He has wenllhere is appropriate, and therefore that lived is to be analyzed as live-en.Th e present participle inflection, which we have labeled -ing, presentsfew difficulties. Aside from the fact that the final "g" !!ound is sometimes

dropped in conversation, it really has no pronunciation variants.The comparative and superlative inflections, which we have labeled

-er and -est respectively, have only a few pronunciation variants. When-er is affixed to the stem good it req ui res a variant of the stem, spelled"bett," to appear, and when -est is affixed, yet another variant, spelled"b ," appears. But as with the other inflectional suffixes, we focus on theabstract and represent the morphemes of better and best as good-er andgood-est respeCtively. Worse and worsl would be analyzed similar ly asbad-er and bad-eJt.

EXERCISES1 All the words .given below begin with a prefix. It is possible to

SQrt them Out and make a list, exactly like the one in 2.14,comai'.l.ing six meaning types with twO prefixes of each eype.(a) Make such a list and then (b) go to a college level dictio'naryand see how many other example words you can find for eachpreqx and add them to your liS{; then (c) identify the root ofevery example word o n you r list and de termine w h e t h ~ r it isbound or free.d ~ e n t r a l i ~ l o y . rrelevantd ode defrost insanesemit'onscious amoral asexualdis,courteous transmit foreshadowtrif ycle disobey supersonicarchduke semiofficial superstructurearchenemy mono/,Jane asymmetricalde-escalate transplam monotheismtransfer protoplasm f o r e t ~ l l semifinal tripod archtraitordislike imtroper prototypeiyogical monorail forewarn


11/12

,>

2

3

ENGLISH MORPHOlOGY

All th e words given below end with a derivational suffix. Icisossible to sort them out on th e basis of the final suffix and makelist similar to [he on e in 2.23 .AIl the function groups of 2.23

bu t the last (ADV < ADJ/N) are represented . However, the::- 1ITl..-5'.. h:iy on e nr tWO and not ruways three suffixes re pDiay ue 0 1 k 23 d h (b) 00senting a given group. (a) Make a list li e 2. ,a n t en c -

fining your analysis to the last suffix only, try to fin? otherwords ending in the same suffix and add them to your ! J ~ t , then(c) identify the root of every example word on your list anddetermine whether it is bound or free.alarmistauthorizationBaptistcapacitatece llist

I certific ation \ChineseChristendomcivilization. comestantcreamydeodorantearldomearthenElizabethanexplosionfacilitate

formalizationfreighterfruityglorificationglorioushyphenateicyIndo nesianinformantinhabitantJapanesejournalesejuicykingdomleaden.".! . Ondonermeaty

orchestrate 'ParisianparticipantpoisonousPortugueseralOYRepublicanriotoussandysilkenstardomstarvation'Steamer_stylistthirstytraitoroustypist

-odificarion vaccinateamousfixation novelist waxenflowers officialdom wooden

,

Printed below is a passage by Henry David T ~ l O r e a u . with ~ o r t y words italicized . Each i talicized word ends ID ~ . n mflecuonalsuffix. Copy the words on a sheet of paper. and beSide .each w ~ r d write the morpheme label for the inflec tional suffix It contatns(-ts , -poss, -prJ, -pst, -en, -ing, -er, -tst). Th e answers fo r the firstfive italicized words are printed below the passage.At a certain season of our lite we dr t dC(uslowud to cOQ.siderevery Spot as the possible site of a house. I ~ a v e thus sucv?edhe coun on every side within a dozen mJles of w ~ e r e live.i m a g i ~ i I have houghl all the farmJ in successIOn. for

O be bouhht and 1 knew their price. I walked over eacwere t /Y ' " d t d' d on hus!armrr's premises, tasuJ'-his wil dPP ~ s , l ~ o u r s e .-bandry with him, look his farm at hiS pnce, at any pnce,

41EXERCISES42

4

5

mortgaging it to him in my mind; even PUt a higher price onit,-100k every thing but a deed of it,-tOok his word for his -deed, for I dearly love to talk,-cultivated it, and him tOO tosome eX tent . I trust. and withdrew when I hdd enjoyed it longenough, leaving him to carry it on. This experience entitledme to be regdrded as a son of real-estate broker by my friendJ.Wherever I Jdt, ~ h e r e I might live. and the landscape radialedfrom :!pe a


12/12(

ENGLISH MORPHOLOGY

We know that the elements between commas above are wordsnot only because they manifest internal stability and positiooalmobility bur also because they require a special kind of pronunciation pattern that has the effect of tying twO roots close to gether. {Notice how especially loud the first syllable is in all thecited words.) Such words are quite common in English and arecalled compounds. There are many types of them. Notice bowthe words in (a) seem to be composed of (W O noun rOOts ; thosein (b) of a noun root, a verb root, and a derivational suffix; thosein (c) of a noun root and a verb root. See if you can think ofother compounds of each of the three types listed. Examine apassage of English prose and see how many compounds it con-tains. look for yet other types besides the three types givenhere.

FO R FURTHER READINGNida ( 1946) treats segmentation and (he identification of morphemes

and presents many problems for practice. Gleason (1961 , pp. 51-127)surveys American s[[ucturalist morphQlogy and discusses many detailsof English. Quick et al. (1972, pp. 973-1008) treat English prefixes andderivational suffIXes in some detai.l; they treat inflectional suffixes in con-siderable detail at various places throughout their work. Marchand(1969) presents the most comprehemive treatinent of English prefixesand derivational suffixes. Joos (1957) includes highly technical articlesabout both general questions of morphology and specific details ofEnglish.

(

43

Date post:	05-Apr-2018
Category:	Documents
Upload:	jose-luis-arriola-pena
View:	227 times
Download:	0 times

38840847 English Morphology

Documents