Part-of-Speech marking in an IAL

This essay will be concurrently posted on my "Views of Language" website and the AUXLANG mailing list. It is the first of a series of essays describing my thoughts on what an IAL ought to be and what it ought not to be. Ultimately it will be followed by other such essays, at least on my web site; if I think there might be interest, they will also be posted on the AUXLANG list.

The subject of marking of parts of speech in an IAL has often come up. Some favor "strict POS marking," which to my knowledge has never been officially defined, but which I understand to mean that all nouns end in one suffix, all verbs in another, etc. (This is not my position, though some have attributed it to me.) Others believe that POS marking is not a good thing, that natural languages do not have it, and that it makes a language look and feel artificial. In fact, mine is a middle position, but closer to strict POS marking than none at all.

To define my position, it is first necessary to consider the distinction between "open classes" and "closed classes." Open classes of words are groups of words (in this context, part-of-speech groups) such as nouns and verbs, with an effectively infinite membership. One can never know all the nouns in a language, because new ones are constantly added. The same for verbs and adjectives. By contrast, close classes are finite, usually rather small groups. It is certainly true that most of us know all (or almost all) the conjunctions of our native language, or the prepositions.

There are two classes that deserve special attention. Adverbs are a very mixed bag. Some, in fact one could say nearly all, are related to adjectives, and since adjectives form an open cless, adverbs must be as well. But there is a small class of adverbs that really ought to be treated separately from the adjective-based ones. They have all, in traditional grammar, been combined into one part of speech, but words like "not," "never," and the like should really be considered as a different class. And yet another group, such as "in," "out," and the like, are really more closely related to prepositions than to anything else. In fact, one can consider them to bear the same relationship to prepositions that intransitive verbs do to transitive verbs. These latter two classes should really be treated as small, closed classes, and the open class status of the adverbs reserved for the adjective-based ones.

The other class to consider is the pronouns. In this case, I believe that exactly the opposite needs to be done from what I have written about the adverbs. They have traditionally been set out as their own closed class. But they function almost identically to the nouns, and it would seem to me that a single class (open, of course!) including nouns and pronouns should be considered for marking purposes. (Of course, new pronouns do not get added very often. But most usages of nouns are of words which have been in the language for centuries.)

Once this classification into open and closed classes has been made, my position on POS marking is easy to state:

The reason for this is that, while most words one runs across are likely to be familiar, unfamiliar ones may appear at any time, and it is important for an IAL user, who is always going to be a second-language user, to have a handle on the syntax. POS marking is not a total solution to the problem of unfamiliar words, but is an important ingredient in promoting comprehension.

Last modified by B. R. Gilson (brg1942@gmail.com) July 5, 2010.



Sign up for your own site

 Back to the essays main page

 Back to my home page





Free web hostingWeb hosting