Basic English Syntax with Exercises

arbitrariness

1.2.1 The Lexicon

It is obvious that this knowledge is not predictable from anything. There is no reason why the object that we call a cat should be called a cat, as witnessed by the fact that other languages do not use this word to refer to the same object (e.g. macska (Hungarian), chat (French), Katze (German), gato (Spanish), quatus (Maltese) kot (Russian), kissa (Finnish), neko (Japanese), mao (Chinese), paka (Swahili)). Moreover, there is nothing about the pronunciation [k?t] that means that it must refer to this object: one can imagine a language in which the word pronounced [k?t] is used for almost anything else. This kind of linguistic knowledge is not ‘rule governed’, but is just arbitrary facts about particular languages.

1.2.2 Categories

Fortunately, there are other ways of determining the category of words, which we will turn to below. But it is important to note that there are two independent issues here. On the one hand is the issue of how the notion of word category is instantiated in the linguistic system and on the other hand is the issue of how we, as linguists, tell the category of any particular word. As to the first issue, word categories are simply properties of lexical elements, listed in the lexical entry for each word, and, as we have pointed out, lexical information is arbitrary. Therefore, word categories are whatever the linguistic system determines them to be. While there may be some link between meaning and category established by the linguistic system, for now it is not important that we establish what this link is or to speculate on its nature (does meaning influence category or does category influence meaning, for example?). More pressing at the moment is the issue of how we determine the category of any given word. Before looking at specific categories, let us consider some general ways for determining categories.

8.2.2 Control

There are a range of possible ways to interpret this sentence depending on who is taken as the antecedent of PRO. The arbitrary reading is possible with the meaning that if anyone were to leave this would make things difficult for you. On the other hand, the sentence might mean that I am the one who could leave, and my leaving would make things difficult for you. Finally, it could mean that you are the person considering leaving and if this happens you will experience difficulties. All these cases involve conditions which are not possible for PRO in other contexts. In the first case, PRO has no antecedent, as already discussed. In the second the referent stands three clauses above the clause of which PRO is the subject and PRO typically has to have its antecedent in the next clause above it. Finally in the third possibility the antecedent is not in a structurally superior position and so it does not even count as a binder. Due to these observations, it has been claimed that in these circumstances PRO is free to refer to any element, as long as the conditions for that element are also respected, e.g. PRO would not be able to refer to a pronominal in a lower position in its own clause as this would give the pronominal a binder within its binding domain, which is not allowed for the pronominal.