Class Discussion, February 28

Transcription

1 Class Discussion, February Morphophonology PHONOLOGY: The component of a grammar made up of the elements and principles that determine how sounds pattern in a language MORPHOLOGY: The system of categories and rules involved in word formation and interpretation. o MORPHEME: the smallest unit of language that carries information about meaning or function MORPHOPHONOLOGY: The study of the interaction of word formation with the sound system of a language. In phonology, we studied the relationship between phonemes vs. allophones. In morphology, we are interested in the relationship between the underlying forms of morphemes and their allomorphs o ALLOMORPH: variants of a morpheme whose distribution is determined by their phonological environment We can write rules describing the relationship between the underlying form of a morpheme and its allomorphs in much the same way that we wrote phonological rules. For rules involving morphemes (segments of sound with meaning attached), don t worry about writing the underlying form and the allomorphs in terms of features. Just write the forms. o But: put the environment in phonological features. The rules that determine the distribution of allomorphs are subject to many of the same considerations as found with phonological rules. o Place and manner of articulation assimilation for ease of pronunciation o Application of processes like epenthesis or deletion to create a particular phonotactic structure allowed by the language. 1

2 1.1 Russian Example [otjexatʲ] [otstupitʲ] [odbrositʲ] to have ridden off to have stepped back to have thrown aside What prefixes do you see in the data set? No minimal pairs in the data set above, but Do we want to say that Russian lacks phonemes /d/, and /t/? Nope. There are minimal pairs elsewhere in the language. [ta] that [da] yes In the ot/od alternation, we are concerned with prefixes that have meaning. They are morphemes. o /t/ and /d/ the phonemes are just sounds. They don t have any meaning attached. Writing a rule: Figure out what the underlying form of the prefix is: [ot] #_j #_s [od] #_b Given the small data set, it is a little hard to tell, but since voicing seems important (that is the difference between [od] and [ot]), is there one form of the morpheme that only occurs in an environment with other phones of some particular value for [voiced]? o [ot]: [j] is [+voiced], [s] is [-voiced] o [od]: [b] is [+voiced] /ot COMPLETED/ has two allomorphs, [od] and [ot] /ot COMPLETED/ [od-] / [+voiced, -continuant] [ot-] / /elsewhere 2

3 1.2 English Plural Morphology This question, like Russian, deals with multiple allomorphs of a single morpheme. [kæts] *[kætz] [lɪps] *[lɪpz] [dɑɡz] *[dɑɡs] [fɪbz] *[fɪbs] [kaʊz] *[kaʊs] [kɑrz] *[kɑrs] Do we want to say that [s] and [z] are allophones of a single morpheme? / z / [s] / (in some environment) [z] / elsewhere No, because we can find minimal pairs in English illustrating that English has phonemes /s/ and /z/. [sɪt] and [zɪt] As a result, we can t say that [s] is really just an allophone of /z/ (or vice versa) However, at the end of a word, we see that the distribution of [s] and [z] is dependent on the phone adjacent to it: if the last phone in the word is [+voiced], use [z]. If the last phone in the word is [-voiced], use [s]. Can we write the rule like this? Let s say that /z/ is the underlying form here. /z/ [z], elsewhere [s] / [-voiced] σ Now, we can t finish here. What does this rule predict? (This rule predicts that the word [sɪt] isn t possible in English it should just surface as [zɪt]. [s] only surfaces when in a coda after a [- voiced] phone. /zɪt/ [zɪt] So, how can we distinguish between /z/ and [s] in the data set above and /z/ and /s/ in words like [zɪt] and [sɪt]? 3

4 The instances of /z/ that we re interested in the data set above have something in common with each other: they carry meaning. They indicate that we re talking about multiple cats or dogs or cows, etc. The sounds [s] and [z] in words like [sɪt] and [zɪt] don t mean anything. o They are just phones in the language, whereas /z/ with its plural meaning is a MORPHEME. MORPHEME: the smallest unit of language that carries information about meaning or function Morphophonology: the interaction of word formation (morphology) with sound systems of languages (phonology). o Morphemes have underlying (or mental) representations, like phonemes. And there are also surface forms called ALLOMORPHS (like ALLOPHONE). We ll write a rule that links meaning to pronunciation. It will tell us where a morpheme with some underlying form (like /z/) is realized with some other form. For rules involving morphemes (segments of sound with meaning attached), don t worry about writing the underlying form and the allomorphs in terms of features. Just write the forms. Put the environment in features. For now, I will just tell that the /z PLURAL/ is the underlying form. We ll get back later to why I think this is true. / z PLURAL / [s PLURAL] / [-voiced] [z PLURAL] / elsewhere We ve already seen that manner of articulation assimilation (voicing) happens with morphemes and their allomorphs (Russian). Other processes that we ve seen happening for phones also happen with morphemes. Epenthesis: [fɪʃɪz] *[fɪʃz] *[fɪʃs] [lɑsɪz] *[lɑsz] *[lɑss] [lɪzɪz] *[lɪzz] What is the configuration of sounds that seems to be violated in the starred words in the far right column? Write a phonotactic constraint. o Two [+strident] sounds can t be adjacent in a coda. 4

5 What form does the plural morpheme take in order to avoid violation of this phonotactic constraint? We can add the form to our list of rules. / z PLURAL / [s PLURAL] / [-voiced] [ɪz PLURAL] / [+strident] [z PLURAL] / elsewhere Why did I pick /z PLURAL/? First piece of evidence: Our coda constraints wouldn t rule out the *-marked (ungrammatical) form below, but my intuitions are that I pronounce it with [-z]. We should make the underlying [-z] because our rules would permit us to keep [kɑls]. [kɑlz] calls *[kɑls] but! [fɑls] false Second piece of evidence: One change to get to [s PLURAL] (voicing) one change to get to [ɪz PLURAL] (epenthesis) What if I picked /s PLURAL/? One change to get to [z PLURAL (voicing) Two changes to get to [ɪz PLURAL] /s/ [ɪs] (epenthesis) [ɪs] [ɪz] (voicing assimilation, where I assume all vowels are [+voiced]) 5

6 2.0 Morphology MORPHEME: smallest unit of language that carries information about meaning or function o SIMPLE WORDS: words that consist of a single morpheme; they can t be divided further o COMPLEX WORDS: words consisting of more than one morpheme; they can be divided further cat = a simple word: one morpheme, [kæt] cats = a complex word [kæt] + [s PLURAL] swanboat = a complex word [swɑn] + [bot] swanboat consists of two of the same type of morphemes o swan and boat are both FREE MORPHEMES: both can be a word can be a word by itself cats consists of two different types of morphemes. o [kæt] is a free morpheme: it can be a word by itself o [s PLURAL] is a bound morpheme: it can t be a word by itself 6