Stephen van Vlack

Sookmyung Women`s University

Graduate School of English

Introduction to Linguistics

Fall 2004

Week 9: Questions for Singleton, Chapter 6

1. What are some different phonological variables and how do they work?

The two main phonological variables are segmental features (phonemes) and suprasegmental features like stress/tones. The basic belief is that these two are the necessary variables or elements for differentiating sounds. The first of these variables, the sounds themselves puts forth the simple claim that words are essentially differentiated by the actual sounds present and the order in which those sounds will appear. Thus, the sounds and their particular order will need to be preserved in the lexicon, but things are not so simple. It is also believed by most generative linguists that sounds themselves are composed of smaller parts or features. This is called distinctive Feature Theory.

First developed by the Prague School of Linguistics in the 1920s, Distinctive Feature Theory claims that each sound is made up a set of specific features for which they each have a value, expressed in the binary marking of either + or -. Each specific sound in a given language is composed of a particular set of these features which differentiate it from all the other sounds in the language. In this view no two sounds in a language have the same exact set of features and values related to those features. To use the generalized term, each sound is composed of a distinctive feature bundle and no two distinctive feature bundles are the same. It should also be noted at this point that the sounds used in language and marked in the lexicon are called phonemes. It is further claimed by Chomsky & Halle (1968) that all speakers of a given language have the same phonemes in their phonemic inventory. The phonemic inventory is simply the total number of distinctive phonemes that a language uses to create meaningful words or lexical units. This seems straight forward enough but there is a further point to be made about phonemes.

There is a bit of a possible problem here in that it is clear that no two native speakers pronounce words in exactly the same way yet they are supposed to all contain the same exact phonemes in their phonemic inventory. Based on this it has been posited that phonemes are idealized representations of sounds upon which the actual pronunciations that emerge from people`s mouths are based. Thus, we can posit two different levels in the process of producing word units. There is the level of the Underlying Representation (UR) which is composed of phonemes and which is the same for all speakers of the language. Then there is he level of the Surface representation. The surface representation (SR) may or may not vary from the (UR), but most often they do in one way or another. The idea is basically that even though all speakers use the same Underlying Representations pronunciations vary. SRs are, therefore, derived from URs in the same way that morphologically complex words are derived form base forms or S-structure is derived from D-structure. Just one interesting thing that relates to what Singleton discusses in this chapter is Chomsky & Halle`s (1968) claim that English orthography is based on Underlying Representations.

The other variable occurs on a different level than the segmental one. Stress and tones exist at the words level (suprasegmental level) to distinguish one word from another. Thus, in English each word of more than one syllable has a particular syllable that receives more stress than any other. Tones in languages like Chinese and Swahili can be said to work in the same way. Different tone patterns on words distinguish words in those languages. Stress works as a way of providing information about a word helping the speaker to distinguish the word form others. There are in fact some pairs of words in English which vary only in their stress placement.

pro`cess (NP) process` (VP)

pro`test (NP) protest` (VP)

de`crease (NP) decrease` (VP)

re`cord (NP) record` (VP)

im`plant (NP) implant` (VP)

The idea that Singleton wants to make here is that the system of stress placement in English is not merely idiosyncratic as has generally been claimed but would also seem to be systematic in one way. Thus stress placement in this group seems to have some sort of motivation. It cannot be a coincidence that in this group of words that share segmental structure for both nouns and verbs that there is a pattern of stress placement. It should also be noted that not all segmentally identical pairs are like this. Some have the same stress pattern, so this is not totally systematic. We might even argue that it is a stress marking process that is undergoing some sort of lexical diffusion.

2. What is lexical diffusion and why is it important?

Lexical Diffusion is the basic idea that structural changes in language whether phonological, morphological, or syntactic occur in stages more or less determined by lexical groups. This idea has come about of observations of language forms over the entire written history of a words existence, taking careful care to notice changes in both form and the structural processes the word can engage in. Inn these investigations by historical linguists it was found that a specific change will first affect a specific, highly limited group of lexical items which seem to have something in common. The change will then eventually spread in stages affecting more and more groups of lexical items until the change has run its course and affected all or the vast majority of the items that it might lay claim to.

If you look at the Old English data which I posted in he answers to Singleton Chapter 3, you can see that there used to be a tremendously varied array of different morphological endings used to mark the plural. Looking at modern English it would seem reasonable to ask the question how then did this `-s` plural marking system develop. Well, if you look at the old English data you can see that masculine nouns has a plural marking -as in the plural nominative and accusative case forms. It can surmised that at some point the `a` in the -as dropped off. The ending then began to spread, presumably first to other masculine nouns and then to neuter nouns, as they have more similar endings. Lastly the -s ending probably spread top feminine nouns as their ending were quite different from those of the masculine and neuter genders. In any case, the important thing to remember is that change does not occur instantly. It takes place in a series of steps and these steps seem to b controlled by the lexicon.

Lexical Diffusion is an important idea for us for two main reasons. The first of these is that it gives evidence to our claim that lexical items are arranged in neural networks, for it would sem obvious from the description given by Singleton (2000) that there is some network working in the initial spread of the `-ing` phenomenon in French. The other important claim that Lexical Diffusion makes is that lexis and lexical organization does indeed seem to have an effect on structural (grammatical) operations. After all, plural marking is a morphological operation. The -s plural marker of modern English might seem systematic and productive to us right now (therefore excluding it from the lexicon entirely in stricter genitive views) but there was an extended period of time in which it was not so. Undoubtably there was a period when the lexicon determined how the spread of the phenomenon would occur and who it would affect next. Undoubtably similar processes of lexical diffusion are taking place in English right now but the processes of language change are slow and generally only recognizable over an longer period of study and comparison.

3. How does orthography relate to lexis?

Orthography is a system developed to represent words and as such is closely connected to lexis itself. Some languages have developed orthographic conventions which relate specifically tho the meanings of words such as capitalizing nouns like German does and English used to do. Also there are ways that some orthographic systems have adopted ways of dealing with homophony. Thus, one of the reasons English has such disparate spelling for similar sounds is the high level of homophony in the language. There are many words which sound the same but are in fact different words which just happen to share the same sound (bare - bear). Some of this is a historical accident where the words in an older form of the language were pronounced differently but came to be pronounced the same due to changes in the language`s sound structure (phonemic inventory). But it bears remembering that spelling conventions are just that conventions and can be changed and often are (take the Americanisms of Daniel Webster for example) when people so desire.

4. How do associations between signs and categories of lexical items work?

Well the simple example would be Japanese hiragana, which is used for writing only morphological endings and kana which is used for foreign words. Each of these work on different levels and types of systems. In German a similar case exists with the letter `c` which is used in non-cluster positions to indicate foreign borrowings. Even in Korean it is easy to see which words come from English due to the proliferation of empenthetic vowels // used to break up consonant clusters not allowed in Korean. These vowels could be claimed to be markers of Englishness in words as they are not really necessary. It would be easier, in fact for Korean to simply simplify the clusters, as in the case of autobike /tobi/ and not //tobik/, but this is a rare case. Usually the consonant cluster is preserved in the number of segments but the segments are broken into smaller units through the insertion of an empenthetic vowel. Thus English loan words have a substantially different and highly marked structure which strongly contrast them with other Korean words. In this way the writing system has chosen to mark English words differently by using a certain vowel rarely used in other Korean words.