Herslund on Danish suprasegmentals: an appraisal and a comparison

Abstract

ABSTRACTMichael Herslund has, as is commonly known, contributed very significantly to Romance and general linguistics, both synchronically and diachronically. But he has also contributed important analyses of his mother tongue, Danish. I have had the privilege to discuss phonological and morphological problems with him, both concerning Romance (particularly French) and Scandinavian (particularly Danish) languages. Here I continue the discussion regarding Danish suprasegmentals, focusing on the entire complex around the Danish laryngeal syllable rhyme prosody called “stød”. In section 2, some basic prosodic features of Danish are discussed, emphasizing Herslund’s contributions and their relation to my own Non-Stød-Model. Section 3 presents challenges posed by Michael Herslund to the Non-Stød Model that have not yet been countered in a wholly satisfactory way. These challenges particularly concern circularity and productivity. This section is just one example of Michael Herslund’s importance for the study of his Danish mother tongue.KEYWORDS: Danishstødmorphologyphonology AcknowledgmentsThis paper is based on several earlier publications of mine (2005, 2008, 2014, 2018, 2019, 2022a, 2022b). I am indebted to Nina Grønnum for many valuable comments on the manuscript (both its first and final version), mainly – but not only – concerning style, and to two anonymous reviewers for numerous insightful comments and suggestions that certainly improved the paper.Disclosure statementNo potential conflict of interest was reported by the author.Notes1 Herslund says in a note that “This is essentially the analysis proposed by Andersen (Citation1954) into ‘α and β words’”. See the next note.2 Here is a conflict with Andersen’s distinctions: Andersen’s “α-words” only correspond to Herslund’s monosyllabic Type A-words, whereas Herslund’s polysyllabic Type A-words correspond to Andersen’s “γ-words”.3 Inflected forms of the latter as naboen (sg. def.) [ˈnæːˌb̥oːˀn̩] and naboer (pl. indef.) [ˈnæːˌb̥oːˀɐ] have long vowel and stød in the second syllable, and those forms thus exhibit the same type of secondary stress as anse.4 This agrees with Martinet’s (Citation1937, not quoted by Herslund) use of the category “i de très faible intensité”.5 A comparison between Herslund’s treatment of syntactic stress and mine is beyond the scope of the present paper; the same applies to his analysis of intonation: section 1.4.4 (Citation2002, 28–30), based upon Grønnum (Citation1992). (The final chapter of Basbøll (Citation2005, 514–541) concerns “From word to utterance”.)6 Cases where words ending in an obviously bimoraic syllable do not have stød, e.g., loanwords like spleen and native words like fersk ‘fresh’ (due to a sound change whereby a voiceless, and thus obstruent, /r/ is turned into a glide, thus a sonorant), are treated as (systematic) exceptions – thus not undergoing any stød-alternations – with the feature {–stød}, see Basbøll (Citation2022a).7 This follows from a subdivision of the vocabulary into nativelike (including Latin, Greek, and German loans) and non-nativelike, see Basbøll, e.g. (Citation2014).8 Grønnum and Basbøll (Citation2001, 239) say that, with three exceptions, “the tendency is for the stød consonant to be slightly longer, and there are 13 instances (from a total of 20) where the stød consonant is significantly longer than the stødless consonant, by 0.8–3.1 cs (table 2)”.9 Herslund gives several valid arguments for his position, most of which are known from Hjelmslev’s (Citation1951) analysis of the Danish expression system.10 The term stages has some dynamic associations, in this respect – not unintentionally – resembling my earlier use of “productivity degrees”.11 The ‘e’ in position P1 for adjectives in my system represents PL/DEF/GRADE, see Basbøll (Citation2005, 365), but this is not relevant for the discussion with Herslund.12 In Basbøll (Citation2005, e.g., 358) I called this default ending “Fully Productive”, but I have now, following some of Herslund’s suggestions, abandoned the terminology of “productivity”, see section 3.13 The PP has the structure [((ud) (ta:l)) t]; the suffix t is in the position P2: it is a stage D suffix, parallel to te in mente (section 2.3.3). There is stød on ta:l since Non-Stød does not apply.14 Udsalg means ‘sales’, so udsalgschef means, literally, ‘manager for sales’.15 Uldall (Citation1936) proposed that the verb stem must be found by subtraction from the infinitive.16 I say “can occupy” because P3 is the canonical position for a default ending, but such an ending is in certain cases lexicalized and will then occur within the basic word (see Basbøll (Citation2005, 367–373)). This problem is dealt with in section 3.3.17 In Basbøll (Citation1998, 73) I said that a typological investigation should decide whether Danish is a typical mora-language, an atypical mora-language, or not a mora-language at all. Obviously, this also depends on the definition of “mora”.18 During language evolution, an ending can become part of the basic word. There are many examples of this natural development.