11 All forms in (18) consist of a stem plus an inflectional suffix [t]. In (18a), the morphologically correct 10 The grammar of diminutive formations involves variation among speakers and gradedness of judgments to a high extent. In this sense, it involves ineffability only for those speakers of German who do not accept any of the options for the diminutive of Monat in (17) (and related cases). 11 It has been claimed that the blocked adjectives denote mental states. However, there are some adjectives that do not denote mental states, like flatt ‘flat’, which are nevertheless subject to the restriction in question. Conversely, glatt ‘happy’, denotes a mental state but is free to appear before a neuter noun (Viktoria Dryselius, p.c.).
Ineffability in Grammar 13 forms would be latt and flatt with a geminate final consonant, which are phonologically fine (compare het-hett ‘hot’), but nevertheless totally unacceptable when used attributively. (18) Gaps in the neuter adjectives in Swedish a. *ett latt barn [l :t/lat]‘a lazy child’ en lat pojke [lat] ‘a lazy boy’ *ett flatt hus [fl :t/flat] ‘a flat house’ en flat tallrik [flat] ‘a flat plate’ b. ett glatt barn[glat] ‘a happy child’ en glad pojke [gl :d] ‘a happy boy’ ett platt hus [plat] ‘a flat house’ en plat tallrik [plat] ‘a flat plate’ ett solitt hus [solit] ‘a solid house’ en solid byggnad [li:d] ‘a solid building’ c. *ett rätt barn[r ε t] ‘a scared child’ en rädd pojke [r ε d:] ‘a scared boy’ Raffelsiefen (2002) suggests that the reason for ineffability in (18a) is that the neuter formation would imply lengthening of the final consonant and a concomittant change in the quality of the low vowel ([a] –> [ :]). This change, being too drastic, is avoided, and no other repair is acceptable. The fact that the change is admissible in [glat]/[gl :d] in (18b) has to do with the independent existence of [gl :d] as the supine of the related verb glädja ‘to make happy’. The phonological form [lat] without gemination violates a minimality condition active in complex formations in Swedish. The ineffability visible in (18c) requires a different explanation. Here, there is a conflict between the voiced (and geminate) ending in rätt (as in en rädd pojke [r ε d:] ‘a scared boy’) and the voiceless inflection t which cannot be resolved. Notice that such a conflict is usually repaired in Swedish, but not in this particular morphological configuration. The explanation for ineffability is to be sought in the forbidden mapping of a particular input onto a specific surface syllable structure. The role of the control component here is to compare a surface syllable structure with an underlying segmental structure and to filter out some configurations in specific morphosyntactic structures.
You've reached the end of your free preview.
Want to read all 40 pages?
- Fall '16