Séminaire de Recherche en Linguistique

Two notes on the morph:eme ratio

De Saussure (1916) defined the linguistic sign as the arbitrary and conventional association of, let’s say, a form and a meaning. Its typical manifestation corresponded for his little brother René as well as for Bloomfield (1933) etc. to Baudouin de Courtenay’s (1895) morfema. With Chomsky (1957) and Baker (1985), the morpheme became the lego of syntax. And with Koopman (2000) and Kayne (2005), the legos became very small and the syntax became very big — such that Starke (2002, 2009) put the syntax into the morpheme. 

In this talk, we’re going to highlight the conundrum the field is presently in, i.e. the need to resolve the tension between tiny morphemes in big trees, and big trees in tiny morphemes, by presenting (I) an argument for the latter and another argument (II) for the former.

(I) A compositional analysis of Mandarin duōshǎo: DM vs Nanosyntax
Mandarin duōshǎo means how much / quanto. However, while English how much clearly has a binary branching structure, made up of degree-how and the positive quantity adjective much, and Italian quanto is (possibly) an opaque portmanteau, Mandarin duōshǎo looks bimorphemic, but the morphemes are not the ones we might expect: (A) the impersonator of "degree-how”, duō, is also the positive quantity adjective ‘much’; (B) shǎo is (elsewhere) the negative quantity adjective meaning  ‘little’. We first attempt several DM-style analyses, noting that they have to compromise either compositional transparency or homophone-freeness. Then we show how a Nanosyntactic analysis manages to avoid the compromises and furthermore establishes a link between the properties (A) and (B).

(II) DemLaDem, or why there’s where and why that’s (like saying that’s) what
In English, (A) where rhymes with there rather than with here, and (B) what rhymes with that better than with this. Leu (2025) proposes that Mandarin suggests that (A) and (B) have the same underlying cause, cause that structurally contains there. In Leu&Tang (submitted), we clarify that, in terms of overt structure, in Mandarin it’s actually there containing that, though that does contain there, in French. We further propose that beyond overt structure — and at the level of UG — the containment is mutual, i.e. (in principle) cyclically recursive. Crucially, that means Demonstratives and Locative adverbs are (or can be) derived from the exact same ingredients, differing in where in the last cycle the derivation stops to spell out. This, in turn, is a rational that motivates segmentation of that and cela into the morphemes we expect to find in definite demonstratives and/or in locatives, including definite articles th-, ce-, l-, as well as locative prepositions at, à, “surviving” as pronounced pieces from an embedded structure.