In terms of its prosodic characteristics, Standard Literary Czech (LC) is fairly straightforward: it has word-initial stress and distinctive vowel length. There are no restrictions on the occurrence of quantity within words or in adjacent syllables in that all types of vowels are found in all positions: čítárna ‘reading room’, dvacátí ‘twentieth’, váhání ‘vacillation’ and robota ‘work’. There are alternations of length both within inflection (e.g., vúz ‘cart’, vozu gen sg) and derivation (e.g., hlas ‘voice’, hlásek, dim), and in some instances the alternation of quantity is a marker of a morphological domain, e.g., infinitival lengthening or genitive plural shortening. So it is curious to find what appear to be very specific restrictions on the distribution of quantity in Czech. I consider one such category, prefixed deverbal nouns, and suggest that Czech does have constraints on the distribution of quantity within a prosodic foot.
Nouns derived from prefixed verbs regularly have vowel lengthening of the prefix, e.g., nahodit ‘to find by chance’, náhoda ‘happening, chance’, načrtat ‘to sketch’, náčrt ‘sketch’, vytahat ‘to extract’, výtah ‘extract’ utratit ‘to lose’, útrata ‘loss’. This is also true of suffixed deverbal nouns as in náčrtek ‘sketch’, výtažek ‘extract’, nadržet ‘to collect’ vs. nádržka ‘reservoir’, vyčesat ‘to comb out’ vs. výčeski ‘hairdo’. There are two main exceptions to prefixal lengthening in this grammatical category: 1) certain prefixes never lengthen, e.g., obklad ‘compress’, podklad ‘foundation’, doklad ‘proof’ vs. základ ‘foundation’ výklad ‘presentation’, príklad ‘example, sample’, and 2) prefixes which ordinarily lengthen sometimes do not, e.g., nadávat ‘to scold’, nadávka ‘scolding’, zacházet ‘to go around’, zacházka ‘detour’, vyplátit ‘to pay out’ vyplátka ‘payment’. The prefixes which do not show lengthening are do-, od-, po-, pod-, pro-, pre-, roz; those which do are: na, pri, u, vy, and za. The reluctance of certain prefixes to lengthen may be due to the consequences of potential lengthening, which in these cases produces more than just a quantitative alternation.
The variation in prefix length of na, pri, u, vy, and za, however, seems to be conditioned by other factors, specifically by the existence of length in the verb root, e.g., nadílet ‘to distribute gifts’, nadílka ‘distribution’, prihlásit ‘to announce’, prihláska ‘announcement’, ukázat ‘to show’, ukázka ‘sample’. These conditions on the occurrence of quantity in Czech suggest that there is a constraint against adjacent long syllables in Czech and that it comes into play only at certain levels of derivation (not so much in inflection or derivation by suffixation, but at the word level derivation). This means that Czech pays a great deal of attention to bisyllabic domains even if this is not immediately obvious. The resolution of a constraint against adjacent long syllables may take the form of shortening either the first or the second syllable, and it is not uncommon for the verb root to shorten in favor of a long prefix: prikázat ‘to assign’, príkaz ‘order’, zachránit ‘to save’, záchrana ‘rescue’, with the result that the two syllable sequence is long-short. The preference for this type of foot over the short-long variant is prosodic and related to stress, whereby the more prominent syllable in the foot is then both stressed and long, a less marked situation than a stressed short syllable followed by a long one. Evidence for such prosodic preferences is given from Spoken Prague Czech, where lengthening is common in contracted forms such as prófa (profesor), véča from večere ‘supper’, ségra from sestra ‘sister’ and in hypocoristics such as Lída for Lidmila, Pét’a for Petr, Rísa for Risard, Bóza for Bozena.