Yes, more or less at this point in the thinking process being followed here, I’ve hit a problem that I’ve been mulling for a while. Don’t be concerned; I’m not going to get personal. It’s, as I see it at the moment, a conceptual problem combined with a nomenclature problem – the two can easily bundle up together into a mess. Initially, it seemed like a difficulty that had become apparent in my approach so far to the issues under discussion; I was even thinking about whether I’d have to rethink everything said up to this point. But now, I’m thinking there is a problem in the perceived wisdom. That’s, of course, an arrogant thing to think but I’m going to pursue it as a real possibility. I’m not going to rethink what I’ve said; I’m going to rethink what others have said.
Analytic and synthetic grammar AND analytic and synthetic languages?!
The problem hit when I had the bright idea of relating the essential grammatical nature and characteristics of English to those of other languages. I started re-reading stuff about language types. The generally perceived wisdom is that there are two key language types: ‘analytic languages’ and ‘synthetic languages’. Some linguists don’t like the term ‘analytic’ in this context and, so, use the term ‘isolating’ instead, so that ‘isolating’ is used in almost direct contrast to ‘synthetic’, which seems conceptually sloppy, as there is no very clear parameter of relationship between the ‘synthetic’ and the ‘isolating’. Conceptually, the contrast is that some languages, the so-called ‘analytic languages’, express grammatical meanings in a generally split-up way between words; a strong example of this is Mandarin Chinese:
Gǒu kàn jī = The dog (Gǒu) watches (kàn) the chicken (jī).
Jī kàn gǒu = The chicken (Jī ) watches (kàn) the dog (gǒu).
Gǒu kàn zhe jī = The dog (Gǒu) watched (kàn zhe) the chicken (jī).
Jī kàn zhe gǒu = The chicken (Jī) watched (kàn zhe) the dog (gǒu).
While other languages, the so-called ‘synthetic languages’, express an array of grammatical meanings within the word; a strong example of this is Latin:
Canis spectat pullum. = The dog (Canis) watches (spectat) the chicken (pullum).
Pullus spectat canem. = The chicken (Pullus) watches (spectat) the dog (canem).
Canis spectavit pullum. = The dog (Canis) watched (spectavit) the chicken (pullum).
Pullus spectavit canem. = The chicken (Pullus) watched (spectavit) the dog (canem).
So, let’s review what’s happening here. On an analytic-grammar level, Chinese shows what’s present and what’s past by adding another word element to show the past (zhe) to the verb (kàn), whereas Latin changes the verb itself from the present (spectat) to the past (spectavit).
Meanwhile, on a synthetic-grammar level, Chinese shows the relationship between the words on the basis of their position. The subject is before the verb and the object after it. The word itself does not change. The word itself does not tell you if it is a subject or an object. These features of its two levels render it, in conventional thinking, an ‘analytic language’, in that grammatical meanings and features are separated between words and not fused into words. Whereas in Latin, the noun words themselves show whether they are subjects or objects by the form of the word: canis and pullus are the subject and canem and pullem are the object. Because, in Latin, the nouns show their functional role in the sentence by their form, we can move the words around in any order in the sentence:
Canis spectat pullum. OR Canis pullum spectat. OR Pullum canis spectat. OR etc.
These features of its two levels render it, in conventional thinking, a ‘synthetic language’, in that grammatical meanings are fused within words: the past is included in the verb by the ‘avit‘ ending and the object nature of the noun (canis) is included in the noun by the ending em in canem.
Trying to articulate the problem
Do you see the problem I’m experiencing? I am making a distinction between analytic and synthetic grammar, which I think works intrinsically, and which expresses the fact that every language has both levels of grammar. But, I am hitting up against a system of nomenclature of language types which has the same nomenclature as my naming of grammar levels. When two distinct but related areas of analysis and description have the same nomenclature, then things get confusing and disarrayed. It won’t do. I feel the distinction between synthetic and analytic grammar is a strong and effective mode of analysis and explanation whereas the distinction between ‘analytic’ and ‘synthetic’ languages is not a useful, or conceptually effective, one.
I feel that the contrast of analytic and synthetic languages is misguided on two grounds and needs to be discarded. YUP! This language-type nomenclature seems almost to claim that a language is either, or predominantly, analytic or synthetic, whereas I am saying that every language has both levels in it, certainly in terms of its grammar, equally. Secondly, this language-type nomenclature confuses ‘analysis’ with the ‘analytic’ and ‘synthesis’ with the ‘synthetic’. ‘Analysis’ and ‘synthesis’ are processes of exploration – in the first, we pull things apart while, in the second, we put things together. But, in the ‘analytic’, we cross-relate things that are alternatives within a vertical hierarchy of alternatives, as discussed previously here; and, in the ‘synthetic’, we perceive relationships between things on a ‘horizontal’ plane of reference. Languages that deal with grammatical-meaning elements in a split-up, separated way can’t in this sense be said to be ‘analytic’; languages that fuse grammatical meaning elements in a fused way within the word can’t be said, in this sense, to be ‘synthetic’. What we are dealing with here in this approach to language types is form features and not processes of reasoning. You might say that a grammatically ‘split-up’ language demonstrates that the grammatical elements are all quite fused in the brain of the speaker before they are expressed, and in the process of speaking or writing those fused grammatical elements undergo analysis into grammatically separate elements – but that’s just in the process of speaking; in the process of listening, they may well be taken from being disparate in the language to being fused together in the understanding consciousness of the listener, but that would be synthesis in the case of a conventionally analytic language.
Trying to work towards a solution
The key thing here is that language types are not being contrasted by their processes of reasoning, insofar as we can judge these at all, but in their form and in how they match grammatical elements of meaning to form: are they like Chinese and have forms that match a single grammatical meaning to each word, or are they like Latin and match multiple grammatical features to each word? So, as we are categorising languages by their meaning-to-form characteristics, it would be better to have nomenclature that does that more expressly. As a clearer alternative to ‘analytic’ v. ‘synthetic’ languages, the contrast might be between ‘fragmentative’ and ‘combinatory’ languages, or ‘unitary’ v. ‘combinatory’ languages or ‘fissional’ v. ‘fusional’ languages.
Interestingly, whatever terms we use for the contrast of language types, English is somewhat halfway between the ends of this spectrum of contrast. Like Chinese, it shows the synthetic structure of the sentence through the position of the words in the sentence (subject first, then verb, then object) but it shows the tense of the verb inside the verb (past as in watched), as Latin does, and not necessarily, though it can (as in did watch), by adding another word to indicate past, as Chinese does.
