Basic concepts

An implicational hierarchy is an abstract object of the following form:

The logic of such a hierarchy in linguistic typology is the following: X is a property that items ai of a language may have. The hierarchy of X is valid for all languages. If, in any particular language, the item ai at a given position of the hierarchy fulfills condition X, then all the items left to ai on the hierarchy fulfill X. The items ai may be any elements or properties that a language may possess. Each ai stands for a proposition applying X to ai. The arrow symbolizes entailment between propositions: if a language has a2, then it has a1; and so forth. Given the logic of implicational generalizations, a1 thus represents the top of the hierarchy.¹ Here are a few examples:

The perception verb hierarchy

Every language possesses a perception sense terminology such that if it has a verb to distinctly designate a sense at some position of the hierarchy, then it has verbs to distinctly designate the senses left to it on the hierarchy. The last verb on the hierarchy will be polysemous with respect to the senses right of it (Viberg 1984).

For instance, Italian has the perception verbs vedere ‘see’ and sentire ‘hear, smell, taste, feel’.

The noun phrase accessibility hierarchy

For each strategy in a language that forms relative clauses, the following holds: if a syntactic function at a given position of the hierarchy can be the function of the empty place in the relative clause (i.e. the virtual function of the head), then all the functions to the left of it can also be functions of the empty place/head in the relative clause (Keenan & Comrie 1977).

A more detailed presentation of this hierarchy is elsewhere.

The plural marking hierarchy

If a language marks plural on nouns of a category at a given position on the empathy hierarchy, then it marks plural on nouns to the left of it (Comrie 1981).

This example shows that the same hierarchy may be operative at various places in a linguistic system. Not only plural marking obeys the empathy hierarchy, but also differentiation of pronominal paradigms and many other traits of a language. Similarly, the noun phrase accessibility hierarchy has been invoked not only for relative clause formation, but also for topicalization and several other syntactic processes.

Foundations

Logically, an implicational hierarchy is a concatenation of a set of implicational generalizations. Methodologically, it is established in the same way as every implicational generalization, viz. by the empirical investigation of variation in a certain domain, classification of the variants and abstraction of the variation principle.

The theoretical interpretation of such a hierarchy is based on the logic of implication: A language has property i at some position of the hierarchy only if it also has i-1. This means that i depends on one more necessary condition than i-1, i.e. it is harder for a language to have. Thus, i-1 is a relatively basic, “normal” thing, whereas i is a relatively complex, “peculiar” thing for a language to have.

For the place of implicational hierarchies in methodology, see the page on exceptions.

¹ The use of the replication arrow ← instead of the implication arrow → is due to this layout requirement.

References

Comrie, Bernard 1981, Language universals and linguistic typology. Syntax and morphology. Oxford: Blackwell (2. ed. 1989).

Keenan, Edward L. & Comrie, Bernard 1977, "Noun phrase accessibility and universal grammar." Linguistic Inquiry 8(1):63-99.

Viberg, Ake 1984, "The verbs of perception: a typological study." Butterworth, Brian & Comrie, Bernard & Dahl, Osten (eds.), Explanations for language universals. Berlin etc.: Mouton (Linguistics 21/1); 123-162.