Spec Tech: Inception
As I mentioned in a post earlier this year (“Language and Culture”, 1/13/11), it is critical to distinguish between what a language has a formalized method for saying and what a language is capable of expressing. Apparently at odds with public opinion, the expressive power of a language is not limited by the words in its lexicon or the structure of its grammar. Despite the fact that the English lexicon does not distinguish between falling snow and snow on the ground (i.e., it is not the case that there is one word for “falling snow” and a different word for “snow on the ground”), we can still talk about it – I just did. The presence or absence of a single term for something (perhaps surprisingly) has little correlation with how important or common it is in a given culture. Just as languages differ in what types of information are encoded in their lexicon, languages also differ in what types of information are encoded in their morphology – what types of prefixes and suffixes they have, and what sorts of meanings are encoded by these word bits. To give a concrete example, all languages have ways of referring to the beginnings of events. However, languages vary widely in how these use their grammar to express this focus (constructions referencing the beginning of an event are typically referred to as inceptives). In this post I’ll survey a few different ways languages express inceptives, and conclude with a little bit about why we should care.
Let’s start with English, since it’s the most widely spoken of the languages I’ll be talking about here. English uses the verbs “start” and “begin” to talk about inceptivity. These are what are called Complement Taking Predicates (CTPs) in English. This means that rather than having just a regular object (like “the cat” in “the dog chased the cat”), the verbs “start” and “begin” take a whole other verb phrase as their object when they are acting as inceptives. We can see an example of this in a sentence like “John started to work”. Immediately we notice something different about the second verb “work”: it doesn’t have any tense marking, like past, present or future. We can’t say “John started worked”, or “John starts works”, or “John will start will work”. This is a common feature of verbal complements across languages.
Next I’ll move on to Navajo, spoken by a little over 170,000 people in the southwestern United States. Navajo uses a prefix to indicate the beginning of an event. The prefix varies based on what word it appears in, but the basic form is something like di- (for the linguists out there, that’s a voiceless unaspirated alveolar stop, not a voiced one). So to express “I started to work”, a Navajo speaker might say “dééshnish” (however, there are other types of inceptives in Navajo in addition to the di- form). In some ways this prefix is similar to English “start”, in that in comes before the main verb in linear order, and the entire verb + verb complex can only have one tense marking. The most superficial difference from English is the fact that the inceptive verb form in Navajo is a single word consisting of a prefix and a verb root, rather than the juxtaposition of two individual verb roots.
Slovenian is spoken by around 2.5 million people, primarily in Slovenia. In Slovenian we find both the English pattern and the Navajo pattern. One inceptive construction is similar to English, in that it consists of a conjugated verb meaning “start” (začel) followed by an infinitive verb. The other construction consists of a conjugated verb with the inceptive prefix za-, as in “zaplaval” (‘he started to swim’). Languages such as this are of interest to linguists because differences between these two patterns can give clues to how speakers organize these constructions.
The final language I’ll look at is Comox, spoken by fewer than 500 people in British Columbia. In Comox, inceptive constructions are formed by a process mostly unknown in English and related languages. The process is partial reduplication, and is executed by copying the first part of a word and adding it to the original word. Comox is part of the Salishan language family, which is well known for its extensive use of reduplication to signal everything from pluractionality (the execution of an event multiple times) to diminutivity (usually the specification of something as small or juvenile). Inceptivity in Comox is indicated by -VC reduplication, which means that the first vowel and the following consonant are copied. Besides this rather radical difference from English in terms of how the construction is formed, there is also a difference in usage. In Comox inceptive reduplication seems to be only used for the beginning of a state, like “get big” (tihih), but not for the beginning of an active event like ‘swim’ or ‘work’.
So why am I talking about a certain type of verb construction in a variety of different languages? It’s useful to know something about the range of variation in human language. We have a tendency to assume that whatever is the case in our own language is the case in every other language. English speakers are likely to assume that past tense is usually indicated by a suffix, that the usual number distinction is singular versus plural, and that inceptive constructions typically involve a set of two verbs, the first of which corresponds to some sort of meaning like “start” or “begin”. However, as I’ve hopefully shown in this post, languages differ rather fundamentally in how they express a certain concept. This doesn’t mean that languages vary in what they can express, but it does mean that expressing a concept is done in very different ways from language to language.