There is a well known theory, widely accepted that as languages evolve their morphological typology changes through the same usual steps.

The major steps are I think isolating or analytic, inflected, and agglutinating, but I forget the exact order and I'm having trouble finding it with an Internet search due to not knowing what it's called, if it even has a name.

(By the way, it's mentioned in James Tauber's answer to "Languages that are gaining morphological distinctions" but isn't named.)

How do linguists usually refer to this trend?

  • Are you sure its 'widely accepted'? The order should be isolating --- agglutinative -- fusional -- polysynthetic, I believe. But I'm really not sure that it works. I'd be interested to see some examples, especially on the polysynthetic to isolating bit, if anyone cared to provide some.
    – kamil-s
    Feb 24, 2012 at 19:57
  • Actually after reading a few questions on this topic here I no longer believe it's widely accepted. It is widely mentioned probably by those that don't know enough about the topic. Also while I'm aware of the forth typology, polysynthetic, I'd never heard it being part of the cycle as far as I can recall. Feb 24, 2012 at 20:03
  • That's a relief, thanks. I'm sure I know examples for fusional to isolating (English) and agglutinative to fusional (Estonian) and maybe I could come up with more if I researched it a bit but I don't think this whole cycle theory can be proven. I think I can recall the first time I heard about it, it was in the context of the 19th c. belief in correspondence between language typology and cultural level, as in Latin with splendid culture being fusional while Indonesian with what they thought was next to no civilization being analytic. This is obviously wrong, in case anyone wondered.
    – kamil-s
    Feb 25, 2012 at 1:17
  • @KamilStachowski: You may be interested in reading the other couple of QAs under the "typological-cycle" tag. Feb 25, 2012 at 10:19
  • 1

2 Answers 2


I've heard of this trend referred to as "cyclic drift" and the "typological cycle," but I'm not sure that there is an "official" name for the hypothesis. After searching briefly, I've come up with a short description and a longer one -- which also cites some good references -- for you. Hope this helps.

  • The link ("a short description") is dead. Jul 3, 2014 at 8:50

Actually I had believed we had fully debunked this theory here some time ago. I thought we had found that the theory was posited by a certain individual linguist but was never accepted. I just had a look around the site looking the name of this linguist, but it seems my memory has deceived me and I can find no such thing.

But I did find what I had not been able to find when I first asked some questions on this topic here. I found a Wikipedia article which covers the topic and includes a list of linguists who contributed to the development of the theory!

The Wikipedia article, Drift (linguistics), contains a section titled Long-term cyclic drift.

Cyclic drift is the mechanism of long-term evolution that changes the functional characteristics of a language over time, such as the reversible drifts from SOV word order to SVO and from synthetic inflection to analytic observable as typological parameters in the syntax of language families and of areal groupings of languages open to investigation over long periods of time. Drift in this sense is not language-specific but universal, a consensus achieved over two decades by universalists of the typological school as well as the generativist, notably by Greenberg (1960, 1963), Cowgill (1963), Wittmann (1969), Hodge (1970), Givón (1971), Lakoff (1972), Vennemann (1975) and Reighard (1978).

  • 1
    I've commented on this elsewhere (eg), but to recapitulate: it was Wilhelm von Humboldt in the mid-19th century who proposed the 'cycle of agglutination' (isolating->agglutinating->synthetic->agglutinating->...), nowadays usually referred to as the cycle of morphological typology. There are numerous problems with this idea and it has not been accepted by linguists in many decades. Drift is a different idea (and BTW that WP article is fairly odd) Oct 7, 2012 at 4:14
  • @GastonÜmlaut, I'm not trying to be nitpicky, I just want to be sure. Do you mean isolating->agglutinating->synthetic->isolating->...?
    – dainichi
    Oct 10, 2012 at 12:52
  • Hi @dainichi, yes you're quite right, that is an error. It does annoy me sometimes that we can't edit comments! Oct 12, 2012 at 0:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.