Quantifying expressions in the history of German: Syntactic reanalysis and morphological change by Dorian Roehrs, Christopher Sapp (review)

The loss of genitive case marking in quantifying expressions throughout the history of German is widely known. While the quantified constituent bears genitive case in older stages of German in cases like 1a, Modern German (MG) requires the quantifier and its dependent to agree with respect to gender, number, and case (1b).

(1)

a. was doch die vrsach sei / das diese vnd dergleichen historien von meniglich wenig beifals vnd glaubens gegeben werde.      (16th c.)¹

  ‘what the reason might be for many people to give these stories little approval and belief’

b. Sie erhielten wenig Beifall/*Beifalls.

  ‘They got little approval.’

Despite the prominence of this change, the book under review provides the first comprehensive account of the history of quantifying expressions in German, based on an impressive corpus with data covering all of the history of German. In particular, Roehrs and Sapp (R&S) look at the development of strong and weak quantifiers and their dependent constituents, including viel ‘much/many’, wenig ‘little/few’, all ‘all’, jeder ‘each’, jemand ‘somebody’, and etwas ‘some/something’. The changes that individual quantifying expressions undergo in the history of German may affect the categorial status of the quantifying word, the form of the dependent constituent, and/or the relationship between the quantifier and its dependent. Though certain changes are shared by all quantifying expressions, other changes affect only some of them. The book’s chapters are hence dedicated to the history of individual quantifying words, including information about etymology as well as the inflectional properties of each quantifier. As regards dependent constituents, a basic distinction is drawn for each quantifier between the cardinal and the proportional constructions, allowing for the generalization that genitive marking in quantifying expressions is restricted to proportional constructions in MG, while in older stages of German dependents of quantifying words may be marked for genitive in cardinal constructions as well; see 1 above.

Ch. 1, ‘Introduction’, highlights some of the differences between the quantifying expressions under investigation, from both a synchronic and a diachronic perspective. R&S argue that these morphosyntactic differences can be captured in a straightforward way when we assume that quantifiers appear in different positions in the functional structure above NP, as a result of different types of reanalysis. In particular, R&S propose four possible stages in the development of quantifying words, illustrated here with the weak quantifier viel ‘much/many’ and the strong quantifier all ‘all’. All quantifiers share the reanalysis from head to specifier (Card >> Spec -CardP), which is unexpected considering current theories of grammaticalization, such as that of van Gelderen 2004. Following Kiparsky (2011), R&S assume that this instance of what they call ‘degrammaticalization’—because of the addition of inflection—is driven by analogy.

(2)

 

The in-depth examination of quantifying words in the history of German starts with Ch. 2, ‘Simplex quantifying word: viel’. Due to the wealth of available data, it is the most comprehensive study of a quantifier in the book and provides the setting for the discussion of the other quantifiers. R&S report that the quantifying word viel itself undergoes a number of inflectional changes, which they interpret as a categorial change from a semi-lexical noun in Old High German (OHG) to a quantifying particle in Middle High German (MHG) to a quantifying adjective in MG. They further show that the cardinal and proportional constructions developed differently. While in OHG the dependent constituents were all marked for genitive, only the proportional construction still allows genitive marking in MG. This development is taken to indicate that only the dependent in the cardinal construction is integrated into the matrix DP due to its smaller structural size (N/A vs. DP), resulting in concord. This analysis also accounts for the word-order change observed for nominal and adjectival dependents, which tended to appear before the quantifier in OHG but...