You are on page 1of 16
Cuapter I. THE EMERGENCE OF MORPHOLOGY. ‘The alm of this chapter Is to introduce morphology. Firstly, morphology will bbe defined and its scope will be celimited in relation to other components of ‘grammar, Next, we will deal witk the emergence of morphology, providing ‘an answer to the questions of when and why this discipline sprouted, as well as who were the most relevant scholars responsible fer its beginning. 1.1. Whar oR WHERE 1s MORPHOLOGY Matthews (1974) defines morphology as “that branch of inguistics which is concemed. withthe forms of words’ in dfferentuses and construction, Morphology can be considered internally and externally, in its relations with other ‘components of grammar. From an internal perspective, morphology can be divided into two main sub-branches: InevecTona, morphology deals with forms of individual lexemes. Tense, number, case, and person are examples of inflectional categories, ‘Woro-ronwanow deals with the formation of new lexemes, which covers the processes cf derivation and compounding. However, the study of morphology cannot be self-contained. The structuraist doctrine Of the rigid separation of linguistic levels is untenable, Although there are some Issues that are the internal concams of morphology, many morphological problems Involve the interaction between morphology and ther modules of the grammar. ‘As regards the Morphology-Phonology interaction, there are some aspects which Clearly betray this relation: (1) The selection of he form theta given morpheme manifests may be influenced by the sounds that realise neighbouring morphemes. Take the indefinite article in English, Ithas two manifestations. itis a before a word that begins with a consonant (@.g. a pear) and an before a word that begins with a vowel (e.g. an orange). We ‘cannot describe the phonological shape of the indefinite article without referring to the sound at the beginning of the word that follows it. 19 {@) Some phonological rules apply to words differently depending on their morphosyntactic properties. For example, some phonological rules are sensitive to the cifference between nouns and verbs (@.g. permit vs per mit), ‘The Morphology-Syntax interaction can be illustrated in the following aspects: (1) Infectional morphology is the fist place where morphology and syntax ‘meet. Inflection is distinguished trom derivation in its relevance to the syntax. The form of a word may be affected by the syntactic construction in which the word is used or by properties of other words in the construction. For example, the choice of the form waik or waiks ofthe verb walk depends on whether the subject ofthe verb is {rd singular person or not (agreement property), the choice of the accusative form of the pronoun depends oniis functioning as the verbal object (coatigurational property). Conversely, syntax needs to have access to morphosyntactic properties. For example, whether a noun is countable or uncountable (inherent property) affects ts behaviour In phrases and sentences (for example, in its combination with determiners: this spade, these spades vs this equipment, “these equipments). Agreement and case ‘are coding properties, which together with behavioral properties are used to define or identity grammatical relations like subject or object (Van Valin & LaPolla 1987: 250). (2) Compounding is seen as being more closely allied with syntax than with xamples ofthe three types of conditioning are given below: 35 ‘Phonetically conditioned allomorph: ural morpherne after voiced segments other than sibilant: 2), ©.9. dogs. after voiceless consonants except sibilant: /s/, ©.9. cats. after siblants (/5.2,)15,3,03): zl, 2.9. boxes. Phonologioally conditioned morphological alterations tend to be very general. Often allomorphs representing different morphemes will cisplay the same phonological alternations if they occur in similar phonological envionment. Thus, for example, the voice assimilation process displayed by the -s plural suffix is not unique to that morphems. The -s mark forthe thd person singular present tense in verbs shows. exactly the same alternations: Present tense 3rd person singular morpheme after voiced segments except siilants: //, e.g. allows. after voiceless consonants other than sibilants:/s/, 6.9. puts. after sibilants ('.2,..3.d3):/zl, e.g. searches. ‘And the same rule applies to English genitives: after voiced segments except siilants: /, e.g. Jane's /deinz/. ater voiceless consonants other than siilants: /), 0.9. MicK’s/miks) alter siblants (/5.2,,1§.3.d3): zl, e.g. Li's Mil. I we make three separate statements, one for the plural morpheme, anather for the third person singular present tense suffixand a third one for the genitive, we miss the {generalization that a sibilant suffix agrees in voicing with the last segment of the {orm to which itis atached, However, this generalisation is captured if we posit just ‘one underlying representation (or base form) for any slant suffix, andifthat underlying representationis converte into diferent phonetic representations by the phonological rules below (@) The underlying representation of any sibilant sufixis /2/ (0) tis realized as: {2 after voiced segments other than sibilants {sl aftr voiceless consonants other than sibllants dial ater sibilants (5,2.5,53,05) ‘This statement shows that the alteration in question is a general phonological process in English, not an idiosyncratic property of any one morpheme. The terms Morpnopioxeuics and Morevorronovogy are used to refer to rules ofthis kind that account forthe realization of phionologically conditioned allomorphs of morphemes. 36 ‘Afurther example of phonological conditioned allomorphs are the diferent morphs representing the past tense morphine: after voiced segments other than fl /a, 6.9 allowed. after voiceless segments other than t: Vl, €.9. blocked. after alveolar plosives (f,d):,i, 2.9. added. We could then posit a generalizaticn accounting for both the sibilant suffix (plural morpheme, present tense singular orpheme and genitive) and the past morpheme: after voiced: fz), fd after voiceless: si, after same sound (i. slllants& affricates: /,2.,\).3,¢5/; alveolar plosives: al: dl, fl More examples of phonologically conditioned allomorphs are provided below: Indefinite atc: before consonants /, ev! before vowels /ar, /ean/ Definite atcte: before consonants /Ba/ before vowels Negative morpheme fn before a labial consonant: fi before a velar consonant: inf before an alveolar consonant or a vowel: in} before fm, n, If: ff before, vi: fg! fn Ln ‘The selection of the alomorph is notrandom. The nasel consonant of the morpheme ins is pronounced aitferently depending on the nature ofthe sound that immediately ‘ollows. “The different allomaorphs of a given morpheme are sald to be in complementary cistsibution, which means that no two of them can occur in identical environments, Solecting one precludes selecting the others. 37 ‘Morphologically concitioned allomorphs: Morphologically-concitioned allomorphs are determined by morshological properties. For example, in sweep and take the presence of the past morpheme determines the ‘choice ofthe allomorphs /swep/ and /tuk/ in swept and took.’ Aitferent example is the comparative morpheme, which takes a different form depending on some ‘morphological properties ofthe adjective: tar, clever-er/ more clever, nteligert erimoreintoligent.® Lexically conditioned allomorphs: ‘An example of a lexicaly-conditioned allomorph is the plural morpheme in oxen? However, Lyons declares that thers tie point in recognizing an allomorph jan of the morpheme (s} in the description of contemporary English since the formation of the word oxen is an irregular fact of English that, despite the segmentabilty of the word, can only be handied by an ad hoc rule applying to this one instance, for ‘although children and brethren have also -en, when segmented, the morphs Nfld! and /bre®/ are notidentical to the morphs representing the singular, whereas /oks/is identical, je. these words have also allomorphy in thei stem. ‘The introduction of the notion of alternant or allomorph carries one important ‘consequence: we can no longer say thet morphemes consist of phonemes, but rather thatthe allomorphs or alternants consist of phonemes. ‘The second problematic case is met when the same form is used to represent different morphemes. There are two diferent cases of this: sometimes the same Phonological form is used to represent different morphemes, but written English ses a different orthographic form to represent each morphemo. This isthe case of outoPoNes. For example, sigh, site and ote are homophones, thati, three different lexemes that happen to be phonclogically identical. Sometimes, the same form (orthographic) is used to represent diferent morphemes, o that different grammatical ‘words are represented by the same word-forms. An example Is the past and past Patticiple of regular verbs. This phenomenon is referred to as Syncaensy, ‘The third case illustrating the lack of a one-to-one correspondence between ‘morphemes and morphs is encountered when a single morph represents different ‘morphemes at he same time, wich is known as Porrwawteau woner (Hockett 1947): 0.9. -8{presenttense}, {Sg} (rd person} was {BE}, {Sg}{Past} "For Hock (087, een example of ormantony meg * For Crystal (1871), oso exams are casos ct phonclegisycetoned alomorhs * Alon nudes ts case among morphological condoned allomerphe 28 Besides, there are cases in whic morphemes are present but are not given any phonetic manifestation, so that tre number of morphemes present exceeds the rhumber of morphs avaliable to represent them, These were callod Zeno sirens by Bloomifeld, and renamed as Zero axoworPus/ ZeA0 woRPHEevEs later: 29. sheep, hit Iwas thought that morphemes might have zero allomorphs, as long as zero could rot be the only allomoroh of a meipheme. For example, we can postulate a zero ‘morpheme for some English pluratnouns like sheep, since itis normal fo plurals to be overtly marked on nouns. However, we cannot set up a zero {singular} morpheme for English nouns, because singularity is never marked by an affix on nouns. But even in the case their postulatn was allowed, they were morphemes without a {ormal realization, which did not fi Bloomfields definition. A diferent solution was proposed for those cases: the plural allomorph might bein the verb form. However, this is conceptually odd, and in ary case the question of whether sheep is plural) singular applies even when the vet's influence is eliminated (The sheep came). ‘The reverse situation can also be found, that is, cases where there is some phonological material thai seems ta belong to:no morpheme at all, so thatthe number ‘of morphs that can be isolatad may exceed the number of morphemes represented. Inthis case, the term Expry wor (Hockett 1947) is used. An example ofthis is the rin chila-r-en: while faild/ and fia! are allomorphs of the same morpheme, and {on/ can be identified with the en! Iroxen, there iso element with which their can bbe compared. Similarly, in sensuay we recognize sens- and -al; however, -u- does rot correspond toa morpheme (cf, persor-al) “The term given to this phenomenon is an unfortunate choice of terminology since if ‘morph is morph by Virtue of representing some morpheme, a surplus word-bullding element that does not represent ary morpheme should not be regarded as a morph. Hence, some scholars prefer the more neutral term Forusnve for referring to any ‘word-bulding element. Sauer (1983) defines tas ‘a distributonal segment of a word- ‘orm segment (ie. an element that recurs in the morphological analysis of word- {orms) independent of whether itis also a morph.” For Bauer (1988:24), this term can be used to be deliberately vague as to whether a particular recurrent form is a realisation of a morpheme or not. We can say then that most formatives are morphs, thats, they represent morphemes, but some are not. They are the 60 called empty morphs. Finally, there are.cases where the representation ofa morphemeis nota segmentable part, where morphemes are not represented by morphs, formal units, but they are processes, instructions to replace one item by another. Those cases are referred to, aS InREGULAR acrERNAOTS, SUBSTITUTION ALTERNANTS OF REPLACE ALLOVORPHS) MORPHS, ‘Some examples are foot-feet, mouse-mice, mar-men, take-took. In cases ike these, 39 Lyons (1968) speaks of ‘words partially determinate with respect to segmentation’ (or indeterminate, in limiting cases ike worse-bad). ‘Ascan be easily seen trom the range of variation the notion of marpheme allows, the Nation of morpheme is not without problems since numerous cases of ‘iltbehaved” ‘morphs are encountered (Anderson 1988), ‘The process of identifying morphemes Is a process not only of identiying single ‘morphemes but of ensuring that when the samme morpheme turns up again, we can recognize itas an example of the same morpheme, i.e. any morpheme should have a fairly consistent form associated with i. But it need not always be precisely the ‘same one, since sometimes a morpheme is represented by very different allomorphs: ‘more ander are both allomorphs of the comparative morpheme. Not only are they phonetically different but one of them is a word (words are composed of morphemes, buthere we have a morpherne one of whose formsis a word!). Could then synonyms be considered allomorphs o the sarne morpheme tao? How much phonetic variation Is tobe allowed between different realizations of a morpheme?, ‘The limiting case of allomorphic variation is found where no generalization can be made in terms of phonological structure or in any other terms about the selection of ‘particular allomorph. In morphology the term SuPrueriowis used for allomorphs of ‘one morpheme which are phonetically unrelated. Examples of this are: qood-better, bbad-worse (instead of “gooder or *badder), and go-went (instead of *goed). ‘Therefore, we are in a position to conclude that the concept af morpheme is enly of limited value. Athough every linguist recognizes the existence of morphemic units in language and view them asthe units thatthe grammatical rules ofthe language have te work on, no linguistic theory these days is based primarily on the morphema Concept due tothe diffcuities of morphemic identification as a whole, 2.3. CLassiFicanion oF MORPHEMES Inthe previous section, morphemes have been studied in connection with related ‘concepts which do not fit the definition of morohemas as segmentable parts which have a formal realization. Some other distinctions can also be drawn, depending on ‘a number of factors, ike their freedom of occurrence in isolation (fae vs bound ‘morphemes), their function or the different effects resulting from their attachment to \Words (inflectional vs dervational morphemes, class-maintaining vs class-changing ‘morphemes), the diferent kind of operations they are involved in (raat, base, stem), the position they occupy in the word (prefixes, suffixes, Infixes), or the different degree of formal modification on the base word (affixes, replacive morphemes, superfixes, and zero affxation). 40 Free vs bound morph (or morpheme) ‘The distinction between free and bound morphs is based on their abilly to appear on their own. A Fase morph is a morph which can occur in Isolation. Free morphemes ‘can be of two different types: Lexical morshemes like cat, which have semantic ‘content, oF function words lke the, which convey grammatical information or logical relations. ‘ABouxo morph is a morph that can only occur in a word form in conjunction with at least one other morph. The morphs dent- and -a in dental are bound, Inflectional ve derivational morphemes" ‘The distinction between inflectional and derivational morphemes is established in terms of a number of diferent criteria, the most basic one being that InrLecTio morphemes result in different wore-forms of the same lexeme, while Derwvariona, morphemes result in new lexemes, belonging to the word-formation domain, Thus, tense, case, person and number are inflectional categories, while the morpheme used fo create negative adjectives, un, is dervational Based on the former two distinctions between bound and free morphemes, on the ‘one hand, and derivational and inflectional morphemes on the other hand, a further distinction can be drawn between GramuanicaL or Fusctional. morphemes and Inevectonat morphemes, on the one hand, and between Deavenona and Lexct, morphemes, onthe other hand. Grammatical morphemes are rea morphemes, since they are functional words like prepositions, conjunctions, articles and pronouns. Inflectional morphemes are bound. Similarly, the opposition oflexical vs derivational morphemes is based on the free/ und distinction: lexical morphemes are free and include nouns, verbs and adjectives, while derivational morphemes are bound, Class-maintaining vs class-changing morphemes, ‘The former distinction is closely related to the citinction betwean Cuass-aananenc and Cvass-cxavaina morphemes, as inflectional morphemes are characterized by being class-maintaining, while derivaional morphemes can be either class-maintaning or class-changing (see Chapter 3) CCiass-maintaining morphemes manly have the function of indicating a particular lexico-semantic characteristic, such as female (6.9. suffix ess), diminutive (0.9 prefix micro- and sufi -ei), collecive (e.g. suffix age), spatio-temporal relations (e.g. prefixes pre- and rans-) and negation (2.9 prefix un), Therefore, they affect the lexical meaning of lexemes and may indicate a subclass within a clas. "The dsineson betwean inaction! and Sains moiphomet wil be addasted in ehapor 3 a Cass-changing morphemes have more abstract meaning and may be seen as markers of syntactic class. ‘There are affixes which are both class-maintaining and class-changing. For example, the suffoc-ly is class-changing in friendly tuming @ noun into an adjective, but class- maintaining in kindly, creating anew adjective from another adjective. Likewise, st is class-changing in traditionalist, a noun derived from an adjective, and class- maintaining in the noun isolationist, derived from another noun. Root, stem, base, affix Morphemes have also been classified In connection with the different kinds of Copefations they are involved in, like whether they are attached to something else or, conversely, get something else attached to them: ‘A Roor is that part ofa word form that remains when all inflectional and derivational suffixes have been removed. Roots can be either tree, like blend in blend-er, or bound, ike dent in dent-l. A Srextis that part ofa word form that remains when all inflectional affixes have been removed. In touch-es, touch is a stem, They differ from roots in thet they can be ‘complex, having elther more than one root, as in redskin-s, ora derivational afix, as in governments. Besides, they are used when dealing wih inflectional marphemes. ‘A.Base is any form to which affixes of any kind can be added. Bases can also be ‘complex: if we add the prefix un- fo fouchablo, we call fouchable a base, a complex Bases and roots cannot be confused as they do not always coincide, that is, they name diferent types of unis: thus, while all roots are bases, not ell bases are roots singe a base can be complex. Hence, a derivationally analysable form to which

You might also like