Saturday, 13 December 2014

A curious case of the "four" (in the Inuit language)

The universe appears to be just one of those things.
-Frank Wilczek

The other day, I was doing an informal presentation on the Inuit language consonant chart to one of my colleagues at the Arctic College in preparation for Inuit Language courses we're simultaneously holding for the next couple of weeks (I'm teaching one of the four classes) when she asked me how I would explain the dialectal variants of the number 4—in some instances, it occurs as "tisamat" and in others as "sitamat".

Rather than giving an unsatisfying "explanation" that it is probably "just one of those things", I decided that I would try looking at it in terms of autosegmental analysis. -I've mentioned "autosegmental phonology" before in this blog which I believe has productive possibilities for the analysis of the Inuit language variations.

Says Wikipedia on autosegmental phonology:

The working hypothesis of autosegmental analysis is that a large part of phonological generalizations can be interpreted as a restructuring or reorganization of the autosegments in a representation. Clear examples of the usefulness of autosegmental analysis came in early work from the detailed study of African tone languages, as well as the study of vowel and nasal harmony systems. (

More specifically, using this gambit:

There are situations in which the rule applies not to a particular value of a feature, but to whatever value the feature has. In these situations, it is necessary to include the presence of the feature, but not to specify its value. (ibid)

I thought some light may be shed on this interesting problem.

How I'd explain the implications of the above insight would be to say that a given word comprises of a series of obligatory slots to which consonants and vowels are assigned. For example, the word for "four" in Inuktitut may be rendered (in highly abstract terms, of course) as comprising of this series of consonants and vowels:

CiCamat—ie, the word is sometimes 'tisamat' and sometimes 'sitamat' depending on how your dialect pronounces the word.

Let me draw your attention to the capitalized "C"s (consonants), where the first consonant is sometimes /t/ and sometimes /s/ while the second consonant slot takes what the first doesn't choose (it comes out as [t] if the first chooses /s/; [s] if the first chooses /t/).

The choice in the expression of the second consonant between the two seems to be psychologically real, and, in fact, it seems to be set in stone: "after thy father, thou may be such but not some other".

I think what is happening here is that the sequence of the first two vowels in the word: the /i/ and the /a/ respectively are exerting some irresistible force in the sequence of this particular slot assignment. I say this because one may have:

sitamat   but not  *satimat


tisamat  but not  *tasimat

In obeying the slot assignments, in fact, one may mispronounce the first two consonants (in this case, change the voicing feature of the consonants in questions while keeping to the vowels as they are) and not lose the meaning:

zidamat  (the hypothetical version of 'sitamat'); and,

dizamat (the hypothetical version of 'tisamat').


Ferdinand de Saussure—and the structuralist school after him—claimed that the production of words in human languages is a completely arbitrary affair:

Saussure posited that linguistic form [ie, a given word] is arbitrary, and therefore that all languages function in a similar fashion. According to Saussure, a language is arbitrary because it is systematic, in that the whole is greater than the sum of its parts.(


There is something incomplete about this perspective. Having seen that /dog/ may express itself as: qimmiq; chien; dog (or, schweinhund as my friend, Kalman, likes to pepper his speech in mimicking a German accent), Saussure makes a rather vague (and odd) claim that "language is arbitrary because it is systematic" and no less "greater than the sum of it parts". He goes from word to language with little or no intermediary steps in between. That he started from word outward apparently gave him no pause for thought, let alone in treating a given word as having no internal structure at all but fully-formed and permanent.

Language structures do have internal logic systems that interplay elemental phono-morpho-syntax with history and linguistic evolution in highly constrained and rigid ways. Rather.

The analysis of "four" in Inuktitut above instead suggests that these systematic linguistic structures are "information-rich" (ie, greater than the sum of their parts) not only because they are capable of encoding meanings (semantic content) but evolve according to lock-stepped iterative processes, much like how plants grow and evolve or how chemical elements combine to make compounds, each according to the available physical properties of its elements and their unique emergent possibilities when combined a certain way.


No comments:

Post a Comment