1. Introduction
It is generally acknowledged that whereas most West Germanic languages have productive diminutive morphology, such as Dutch -(t)je and High German -chen/-lein, the modern Scandinavian languages do not (see e.g. Faarlund, Lie & Vannebo Reference Faarlund, Lie and Ivar Vannebo1997:113 or Alexiadou & Lohndal Reference Alexiadou and Lohndal2023 on Norwegian, Olofsson Reference Olofsson, Grandi and Körtvélyessy2015 on Swedish, or Grandi Reference Grandi2011 for a comparison of Indo-European languages). In this paper we analyse the system of diminutive formation in Viöl Danish, a now extinct Scandinavian dialect which had a long contact history with Low German and in some respects differed substantially from the modern Scandinavian languages. We show that the two productive diminutive constructions in this dialect – one involving the suffix -kǝn, the other a change from common to neuter gender – both had the core meaning ‘small’, but differed in their semantic extensions and pragmatic effects. We argue that while this diminutive system arose partly due to contact with Low German, it had developed its own dynamics in Viöl Danish and appears to have been fully productive when the dialect was recorded in the late nineteenth and early twentieth centuries.
The paper is structured as follows. In Section 2 we define the term ‘diminutive’ and provide an overview of some of the most important literature on this phenomenon. Section 3 introduces the Viöl Danish dialect and the material used for the analysis. In Section 4 we present our analysis of the two diminutive constructions, focusing in particular on the semantic and pragmatic differences between them. This is followed by a discussion of the likely development of the system in Section 5 and a brief conclusion in Section 6.Footnote 1
2. Diminutives
2.1 Definition and classification
According to one well-known glossary of grammatical terms, a diminutive is ‘[a] derivational affix which may be added to a word to express a notion of small size’ (Trask Reference Trask1993:82). In the recent linguistic literature, diminutives have often been discussed under the more general heading of evaluative morphology, i.e. morphology which ‘has the function of assigning a value which is different from that of the standard or default’ (Grandi & Körtvélyessy Reference Grandi, Körtvélyessy, Grandi and Körtvélyessy2015:13; see also Ponsonnet Reference Ponsonnet2018:18–19). Grandi & Körtvélyessy distinguish between four main functional types of evaluative morphemes: the ‘descriptive’ (or ‘quantitative’) types diminutive and augmentative, which prototypically add the semantic values ‘small’ and ‘big’, respectively, and the ‘qualitative’ types pejorative and ameliorative, which express the speaker’s judgement (simply put, whether the referent is considered ‘bad’ or ‘good’).
Diminutives prototypically express small size, i.e. a smaller referent than the referent of the base noun, as in the examples with the Pite Saami diminutive suffix -tj in (1).
However, in many languages, diminutives also have other functions, such as similarity or individuation, which may be more or less transparently related to the notion of smallness. Some of these ‘extensions’ of diminutives in fact belong in the qualitative subtype of evaluative meaning, as they express the speaker’s judgement or attitude, e.g. endearment, approval, or contempt (see Ponsonnet Reference Ponsonnet2018 for a detailed discussion of such extensions). In this study, we follow the definition of ‘diminutive’ used by Jurafsky (Reference Jurafsky1996:534): ‘any morphological device which means at least “small”’, with the addition that the morphological device in question must be productive. In other words, while the same evaluative marker may have many different meanings in a language, for it to be considered a diminutive it must at least have ‘small’ as one of these.Footnote 2 Also, we follow Jurafsky in restricting the term to morphological operations (including changes to noun class or gender); we do not think it is useful to extend the term ‘diminutive’ to lexical devices like adjectives meaning ‘small’, as has occasionally been done in the literature (e.g. Schneider Reference Schneider2003, Gorzycka Reference Gorzycka2020; see also below on Møller Reference Møller1943).
Jurafsky (Reference Jurafsky1993, Reference Jurafsky1996) investigates the various semantic extensions of diminutive markers and provides a useful classification and a semantic map of these. We will only briefly introduce those functions which are relevant to our investigation. An overview of our terminology in comparison with Jurafsky’s is given in Table 1. The semantics and pragmatics of the Viöl Danish diminutive constructions will be explored in more detail in Section 4.
Classification of diminutives used in this study
As Jurafsky points out, diminutives are often linked to the semantic field of children and young animals, and may be used to derive ‘offspring’ nouns in many languages:Footnote 3
We refer to this as the younger referent function of diminutives. Closely related to this meaning is the extension to expressing affection and intimacy, i.e. positive speaker attitude. As briefly mentioned above, however, diminutives may also be used to express contempt and disapproval in many languages. We call these functions hypocoristic and pejorative, respectively. Among the languages surveyed by Ponsonnet (Reference Ponsonnet2018), Ḥassāniyya Arabic is one where diminutives may express a whole range of positive and negative attitudes, as discussed in depth by Taine-Cheikh (Reference Taine-Cheikh2018):
A rather different extension of the core meaning is the one called ‘partitive’ or ‘individuating’ by Jurafsky (Reference Jurafsky1996), where the diminutive is used to refer to a part of a larger whole. Following Rhodes (Reference Rhodes and Tsohatzidis1990:153–154), we think that two subtypes of such ‘partitive’ diminutives may be distinguished, which we will refer to as meronymic and singulative diminutives. In meronymic diminutives, the derived lexeme refers to a constituent part of a larger entity, as in the Eastern Ojibwa examples in (4).Footnote 4
By contrast, in singulative diminutives the derived lexeme refers to a piece or unit of a substance or group. Such diminutives may derive count nouns from mass nouns, as in the Dutch examples in (5), or they may be used to derive individual-referring nouns from collective nouns; see the Kabyle examples in (6).
Two closely related functions are termed ‘small type’ and ‘imitation’ by Jurafsky (Reference Jurafsky1996). In the one termed ‘small type’, the diminutive form does not just refer to a small version of the referent of the base word, but another type of entity which is not only smaller in size but also has a different function or utility. We adopt the term classificatory diminutive from Rhodes (Reference Rhodes and Tsohatzidis1990). Two examples from Rhodes’ work on Eastern Ojibwa are given in (7).
As a further extension from this function, diminutives in some languages express imitation or similarity to the referent of the base noun, as in the examples in (8). We call this function similative.
2.2 Diminutives in Germanic
Diminutives are well attested in the world’s languages. In the survey of diminutives made for the Grambank project, at least one diminutive formation was identified in 661 languages, as opposed to 813 languages without diminutives, i.e. close to 45% of the languages had a diminutive.Footnote 5 Within the modern Germanic languages, productive diminutive systems are found in most of continental West Germanic, though their frequency, functions, and pragmatic effects differ between languages and dialects. Languages with productive diminutive systems include Dutch (Bakema Reference Bakema1998, van der Wouden & Booij Reference van der Wouden and Booij2015, Booij Reference Booij2022b), Afrikaans (Donaldson Reference Donaldson1993:87–95, Coetzee & Kruger Reference Coetzee and Kruger2004), West Frisian (Hoekstra Reference Hoekstra1986, Dyk & de Vries n.d.), Luxembourgisch (Gilles Reference Gilles, Grandi and Körtvélyessy2015), and Yiddish (Birnbaum Reference Birnbaum1979:238–240, Jacobs Reference Jacobs, Rauch and Carr1995). Within the large German dialect continuum – understood here as all varieties that have standard High German as their Dachsprache – there is significant variation, both in terms of the available forms and their usage (Wrede Reference Wrede1908, Ettinger Reference Ettinger1974:47–83, Tiefenbach Reference Tiefenbach1987). Diminutives are said to be more frequently used in ‘Southern Germany, Switzerland, and Austria’ than in the north of the German-speaking area (Dressler & Barbaresi Reference Dressler and Barbaresi1994:103). In fact, in the Low German dialects of Schleswig-Holstein, i.e. the ones historically in contact with Viöl Danish, diminutives have been reported to be entirely unproductive (Mensing Reference Mensing1927–35, s.v. -kǝn). We will return to this point in Section 5.
In contrast to most of continental West Germanic, there seem to be no productive diminutives in the modern continental Scandinavian languages. The most comprehensive Swedish and Danish reference grammars (Teleman, Hellberg & Andersson Reference Teleman, Hellberg and Andersson1999, Hansen & Heltoft Reference Hansen and Heltoft2011) do not mention diminutives at all, and Faarlund, Lie & Vannebo (Reference Faarlund, Lie and Ivar Vannebo1997:113) explicitly state that Norwegian has no productive diminutive formation (see also Olofsson Reference Olofsson, Grandi and Körtvélyessy2015 on Swedish, and Alexiadou & Lohndal Reference Alexiadou and Lohndal2023 on Norwegian). In a number of cases, some constructions have been discussed as diminutives or at least diminutive-like, but none of these satisfy the definitional criteria we use here. For instance, Rosenberg (Reference Rosenberg, Manova, Grestenberger and Korecky-Kröll2024) investigates the Swedish suffix -is, as in cigg-is ‘cigarette’ or bäst-is ‘best friend’, and describes it as having ‘many diminutive-like properties’ (Rosenberg Reference Rosenberg, Manova, Grestenberger and Korecky-Kröll2024:175). However, the function of this suffix appears to be mainly one of amelioration, i.e. expressing positive attitude, rather than expressing small size, and we are not aware of any examples where the coded meaning of [X-is] is clearly ‘small x’ (rather than ‘cute x’, ‘nice x’, etc.). Thus, while we agree with Rosenberg that -is is an evaluative morpheme, it is not a diminutive according to the definition we follow here. For the same reason, the affectionate Norwegian nicknames like Marianne-min and Tone-mor discussed by Kinn (Reference Kinn2020:141–142) do not qualify as diminutives under our definition.
For modern colloquial Norwegian, Fløgstad & Eiesland (Reference Fløgstad and Anne Eiesland2019) have described a potential diminutive construction consisting of the addition of the adjective lit(t)a ‘little, small’ and a change to feminine grammatical gender, as in ei litta hus ‘a little house’ (vs. underived et hus ‘a house’ with the neuter article). The construction may both express small size and a number of the related meanings surveyed by Jurafsky (Reference Jurafsky1993, Reference Jurafsky1996), e.g. affection, contempt, and as a hedging device. Fløgstad & Eiesland (Reference Fløgstad and Anne Eiesland2019) discuss the construction as a whole as a diminutive formation, but it is doubtful whether it can be considered a morphological diminutive in our terms; because the construction does not occur without the adjective lit(t)a ‘little, small’ (see Fløgstad & Eiesland Reference Fløgstad and Anne Eiesland2019:78), one might argue that the meaning ‘small’ comes from the adjective alone and that the gender shift has a different, pragmatic, function (à la those discussed by Steriopolo Reference Steriopolo2021). Hence, while the construction clearly has many similarities to diminutives as understood by Jurafsky, it cannot straightforwardly be characterized as a morphological diminutive.
Finally, diminutives have been argued to be a productive category in (early twentieth-century) Danish by Møller (Reference Møller1943). However, none of the constructions discussed by Møller satisfy our criteria. The suffix -s(e), as in mandse ‘little man’, is better characterized as ameliorative (hypocoristic) and is only very marginally productive, if at all, in modern Danish; the suffix -(l)ing, as in gæsling ‘gosling’ and pusling ‘toddler; darling’, is not productive in the modern language, as Møller (Reference Møller1943:89) in fact indicates himself; and while Møller considers noun phrases with lille ‘little, small’ a kind of periphrastic diminutive, we exclude these on formal grounds. A similar conclusion to ours is reached by Farø & Hansen (Reference Farø and Schoonderbeek Hansen2009, Reference Farø and Schoonderbeek Hansen2010), who report on a short study comparing German, Dutch, and Danish parallel texts. They characterize the suffixes discussed by Møller as ‘more likely relics of an obsolete system’ (2009:148, our translation) and find that translations from German and Dutch usually do not attempt to render diminutives in Danish at all; for instance, the Dutch diminutive huisje is simply translated hus ‘house’. Their conclusion is that ‘there are several qualitative and quantitative arguments for not regarding Danish as a diminutive language’ (Farø & Hansen Reference Farø and Schoonderbeek Hansen2010:11, our translation). By contrast, however, we argue that the Viöl Danish dialect had two productive morphological diminutive constructions which clearly satisfy Jurafsky’s definitional criteria, and which set the dialect apart from the modern Scandinavian languages. In the following section, we give a brief introduction to this dialect.
3. Viöl Danish
The variety under investigation in this paper was spoken in southern Schleswig until the first half of the twentieth century. It belongs to the Jutlandic (or Western Danish) dialect continuum but differs significantly from other Jutlandic dialects, partly as a result of its peripheral position and partly due to its more intense contact with Low German. Viöl (Danish Fjolde) lies in the historical Duchy of Schleswig, which was a Danish fief until it was annexed by the German Confederation in the Second Schleswig War in 1864. Schleswig has been a multiethnic and multilingual polity since the early Middle Ages, the inhabitants speaking various varieties of (Low and High) German, Danish, and North Frisian. The question of its political status was eventually settled with two plebiscites in 1920, which resulted in the northern half joining Denmark (as Sønderjylland ‘South Jutland’) and the southern half remaining within the German state of Schleswig-Holstein.
Viöl Danish was among the Danish dialects which, as it were, ended up to the south of the Danish–German border in 1920. A gradual language shift towards Low German had already been ongoing in this area since the early nineteenth century (see e.g. Bjerrum Reference Bjerrum1990, Pedersen Reference Pedersen2000), and the last remaining speakers of Viöl Danish probably died in the 1930s or 1940s. Fortunately, however, the dialect is comparatively well documented. As a peripheral dialect which had preserved a number of salient grammatical and phonological features from Old East Norse, it had a particular appeal to early historical linguists and philologists. The first linguistic documentation of Viöl Danish was collected by K. J. Lyngby in 1858–59 (on which see Bjerrum Reference Bjerrum1944:8–11). Four translations of Wenker’s German dialect questionnaire were made in the 1880s (published by Ringgaard Reference Ringgaard1964; for discussion see Höder & Winter Reference Höder, Winter, Fleischer, Lameli, Schiller and Szucsich2020, Höder Reference Höder2022). The most substantial documentation, however, was collected by Anders Bjerrum in the 1930s from some of the last speakers, most importantly Catharina (Tinne) Carstensen (1841–1937). Bjerrum published a number of texts with linguistic commentary, a study of personal names, a phonological description (A. Bjerrum Reference Bjerrum1931, Reference Bjerrum1934, Reference Bjerrum1944), and later a dictionary in collaboration with his wife Marie Bjerrum (M. Bjerrum & A. Bjerrum Reference Bjerrum1974). The latter also published a study of participle constructions, a small collection of texts, and a paper on etymology (M. Bjerrum Reference Bjerrum1974, Reference Bjerrum1975, Reference Bjerrum1999). Other studies of individual features include Veirup (Reference Veirup1958) and Rasmussen (Reference Rasmussen1972). There has been little new research into the dialect since the 1970s, however, and despite its interest for the history of Schleswig and for Scandinavian–West Germanic language contact more generally, it appears to be little known outside of Danish dialectology.
The existence of diminutives in Viöl Danish was already noted by Bjerrum (Reference Bjerrum1931:124–125) and mentioned in passing in the dictionary by Bjerrum & Bjerrum (Reference Bjerrum and Bjerrum1974:20, 140, 289). Bjerrum (Reference Bjerrum1934:40–41, Reference Bjerrum1944:222–226) discusses the morphophonology of the diminutive suffix -kǝn, and Höder & Winter (Reference Höder, Winter, Fleischer, Lameli, Schiller and Szucsich2020:82–83) briefly consider this construction in an evaluation of the Danish Wenker questionnaires from Schleswig.Footnote 6 The Viöl Danish diminutives are thus by no means a new discovery, but as far as we know, their meaning and usage have not yet been investigated in any detail.
For our investigation we manually identified all potential diminutives in the dictionary by Bjerrum & Bjerrum (Reference Bjerrum and Bjerrum1974) and the published texts (Bjerrum Reference Bjerrum1931, Bjerrum Reference Bjerrum1975, Ringgaard Reference Ringgaard1964; see also Gregersen Reference Gregersen2025), namely all nouns with the suffix -kǝn (or one of its allomorphs; see Section 4.1) and all nouns attested with both common and neuter gender.Footnote 7 For a number of nouns with variable gender, no semantic difference is recorded. For instance, kas ‘box’ is said to have either common or neuter gender, but there is no evidence of a semantic difference between them. In other cases, there is a semantic difference which cannot be straightforwardly related to diminution. An example of this is tæstǝrmænt, which is neuter in the meaning ‘(biblical) testament’ and common when it means ‘last will’. Such examples, which are possibly better considered separate lexemes, were excluded from the data.
Before presenting our analysis of the diminutive system, we will provide a very brief grammatical overview of Viöl Danish.Footnote 8 Like other mainland Scandinavian varieties, the dialect is predominantly analytic with relatively few inflectional categories. Nouns are inflected for number only, adjectives and determiners are inflected for definiteness, number, and gender, and verbs are inflected for number, person, tense (present vs. past), and mood (indicative vs. imperative); further tense/mood/aspect values are expressed with particles or auxiliary constructions. Personal pronouns have a nominative–oblique distinction, and the genitive suffix -s can be added to personal names and is found in fixed expressions like at oars tii ‘a year’s time’ and te stajs ‘to town’; otherwise case is not a relevant category. Word order in declarative clauses is V2, the finite verb appearing in the second position. Non-finite verbs typically appear before their objects, but a restricted verb-final pattern is also found, most importantly in some uses of the perfect (see Bjerrum Reference Bjerrum1974), as in (9).
As already mentioned above, nouns in Viöl Danish belong to one of two genders, which we refer to with the traditional and more or less arbitrary terms ‘common’ (c) and ‘neuter’ (n). Gender is reflected in several agreeing forms both within and outside the noun phrase: the indefinite article (common an vs. neuter at), determiners, the numerals one to three, many adjectives, and some pronouns. Example (10) shows agreement in two noun phrases, headed by common noal and neuter øj, respectively. Gender is reflected on the indefinite article an/at and the adjective stuort.
The definite article ar does not agree, meaning that many nouns occurring primarily in definite contexts are of uncertain gender, for instance many nouns referring to diseases, like ar febǝrs ‘(the) fever’ and ar kenkhuost ‘(the) whooping cough’. Similarly, because the distinction between common and neuter gender is usually not visible in plural contexts – except on the numerals ‘two’ and ‘three’ – most pluralia tantum are of uncertain gender (e.g. bøksǝr ‘trousers’ and møøpǝl ‘furniture’).
In the singular demonstrative-anaphoric pronoun there is a distinction between common dænd and neuter der. In addition, human beings and some animals may be referred to with the personal pronouns hand (masculine) and hun (feminine), though this is not obligatory, as shown by examples like (11) with the demonstrative-anaphoric pronoun dænd.
Gender assignment is sometimes semantically and sometimes morphologically motivated. For animate nouns, some almost exceptionless rules can be given: words for children and baby animals are neuter, whereas nouns referring to adult humans have common gender (unless they are diminutives). Nouns referring to adult animals may be either common or neuter, though common gender appears to be more frequent. There is thus an association between youth and neuter gender in animate nouns. Inanimate nouns may belong to either of the two grammatical genders, but certain derivational forms are predictable, such as deverbal nouns in -en (always neuter) and deadjectival nouns in -hed (always common). Diminutives are always neuter, no matter whether they have animate or inanimate referents. Collective nouns like fålǝk ‘people’ and fåvltøj ‘fowl, wild birds’, which refer to groups rather than individuals, are also always neuter.
4. Diminutive constructions in Viöl Danish
4.1 Derivation
The Viöl Danish material contains three potential diminutive formations: the suffixes -(ǝ)k and -kǝn, and the aforementioned gender shift from common to neuter. However, the suffix -(ǝ)k only appears in a small number of Low German loan words (e.g. rålǝk ‘yarrow’ and majk ‘cheese mite’) and does not seem to have been productive at all, at least not when the dialect was documented.Footnote 9 For this reason, we will focus only on the two productive formations in the following. We identified 29 gender-shift diminutives and 76 -kǝn diminutives in the material, i.e. 105 diminutive forms in total. Note, however, that the two derivations are occasionally attested with the same base word (e.g. fåvl ‘bird’ and myø ‘young woman, girl’). On the other hand, the base words of 21 -kǝn diminutives are not attested, whether by accident or because they did not exist in the dialect.Footnote 10 The semantic analysis is based on the 84 diminutive forms where the base word could be identified.
The derivation of the gender-shift diminutive is the simpler of the two. It involves only a change of the grammatical gender of a noun from common to neuter, as illustrated by the rule in (12).Footnote 11
No other formal change is involved in the gender-shift diminutive, which we consider a case of zero-derivation from one noun class (gender) to another (see Jurafsky Reference Jurafsky1996:534). A few initial examples of gender-shift diminutives are given in (13).
In a few cases, the simplex word is only attested with common gender, but the neuter is found in one or more compounds which clearly refer to smaller entities, such as the ones in (14).
Note that boam ‘tree’ is attested in several compounds referring to different kinds of trees, all of which have common gender (e.g. danboam ‘fir’, plomboam ‘plum tree’); the only attested neuter compound is pontjǝrboam, which refers to a pole used to secure a load of hay (German Wiesbaum).
The derivation of the suffixal diminutive involves the addition of the morpheme -kǝn to the stem of a noun, following the rule in (15) and illustrated by three examples in (16).
However, the full picture is slightly more complicated than the basic rule in (15) suggests. Firstly, the diminutive suffix has at least three phonologically conditioned variants; after /nʲ/ and /lʲ/ (orthographic <nd> and <ld>) it is normally realized /tjǝn/, after /t/ it is normally realized /jǝn/, and after /s/ it is realized /xǝn/, as shown in (17) and (18).Footnote 12
Secondly, a number of nouns have unpredictable diminutive forms, usually involving shortening or mutation of the stem vowel (see 19) or the insertion of /ǝ/ before the diminutive suffix (see 20). Some are even more irregular, such as the diminutives of vindǝng ‘window’ and tjaab ‘chin’ shown in (21), with loss of /ŋ/ and unexpected use of the allomorph /tjǝn/, respectively.
Most attested plurals of suffixal diminutives are formed by dropping the final /n/ and adding the plural suffix /(ǝ)r/, such as möjkǝr ‘young girls’, lukəkər ‘small trapdoors, hatches’, and hierǝkǝr in (20). A small number, mainly Low German loanwords, have plural forms involving the addition of the suffix /s/, such as møtǝkǝrs from møtǝkǝn ‘nut (fastener)’ and kofərgəns or kofərgərs from kofərgən ‘small chest, box’.
The material contains no examples of diminutives formed from other parts of speech, similar to Dutch examples like kleintje ‘little one, child’ (from the adjective klein) and uitje ‘trip, excursion’ (from the adposition uit) (see Booij Reference Booij2022a for further examples). In other words, all attested diminutives in Viöl Danish have a nominal base. As far as we can tell, the two diminutive formations were both productive when the dialect was recorded in the late nineteenth and early twentieth centuries. Bjerrum & Bjerrum (Reference Bjerrum and Bjerrum1974:16, 20) characterize them as such (see also Bjerrum Reference Bjerrum1931:124–125), and both constructions are found not only with Low German loanwords, but also with inherited Scandinavian words, such as muos ‘bog’, myø ‘young woman, girl’, uost ‘cheese’, and vindǝng ‘window’.
Before presenting the semantic and pragmatic analysis of the two diminutive constructions, it is worth highlighting a few distributional differences between them. Firstly, while -kǝn diminutives may be formed from both common and neuter nouns, by necessity the gender-shift diminutive can only be formed from common nouns. The pool of potential base words for the gender-shift diminutive is thus smaller than for the -kǝn diminutive. Furthermore, while we found a number of diminutiva tantum with -kǝn in the material, i.e. -kǝn diminutives with no attested base word, gender-shift diminutives can only be identified as such if the corresponding common gender form is recorded as well. Finally, we have not found any clear examples of plural gender-shift diminutives. Given that the distinction between common and neuter gender is largely neutralized in the plural, it is possible that gender-shift diminutives simply could not be recognized as such in the plural. By contrast, the plural forms of many -kǝn diminutives are attested, and for a few of them we only have the plural form, not the singular (see hierǝkǝr in [20] above). The fact that we identified more -kǝn forms than gender-shift diminutives is thus unsurprising and cannot tell us anything about the relative productivity of the constructions.
4.2 Meaning and use
Based on the semantic descriptions and language use in the material, we found that the -kǝn and the gender-shift diminutive can both be used with a range of functions. Inspired by Jurafsky (Reference Jurafsky1993, Reference Jurafsky1996) and several of the contributions to Ponsonnet & Vuillermet (Reference Ponsonnet and Vuillermet2018), we decided to map these different functions to the extent possible and see which similarities and differences might be found between the two constructions. The semantic classification was introduced in Section 2, but will also be illustrated with Viöl Danish examples in the following.
As already mentioned in the previous section, the two constructions could both be used with the ‘core’ diminutive meaning smaller referent, where the derived noun simply refers to a smaller version of the base referent:
According to Bjerrum & Bjerrum, shifting the gender of fåvl from common to neuter results in the meaning ‘small bird’. Similarly, adding -kǝn to the noun botǝl results in the meaning ‘small bottle’. In some examples in the material, the smaller referent meaning is supported by an adjective with the same meaning, i.e. there may be semantic harmony between the diminutive construction and an attributive adjective. Compare the co-occurrence of smoa ‘small’ and the diminutive form byntǝkǝr in (23a) with the use of stuor ‘big’ with the base word bynd in (23b).
As discussed in Section 2, diminutives often have functions other than just expressing small size. In the Viöl Danish material, we have found diminutives of both types with four semantic functions: younger referent, classificatory, similative, and meronymic. An example of each construction with the younger referent meaning is shown in (24).
The lexeme myø with common gender refers to a young woman, whereas myø with neuter gender is reported to refer to a young girl, typically of school age. In a similar way, kvej and kvejkǝn refer to a heifer at different ages, kvejkǝn before its first pregnancy, kvej after it has mated for the first time (after giving birth, it becomes a kuo ‘cow’). In these cases, the diminutive thus refers to a younger version of the base referent.
Diminutives with a classificatory function are frequently attested in the material and are found with both constructions, as shown in (25).
The common noun tut refers to a bag used e.g. for flour, coffee, or oats. The gender-shifted diminutive form tut refers not to a smaller exemplar of such a bag, but to a specific type of pointed paper bag used for sweets. Similarly, the derived diminutive huskən does not refer to any small house, but to a specific kind of sod house used for shelter.
The similative type involves referents with are similar to the referent of the base noun. For instance, the examples in (26) refer to children by highlighting a similarity with a physical object (in terms of shape) and an animal (in terms of behaviour).
Note that while these examples obviously refer to children, they do not fit neatly in the category younger referent because the referent of the diminutive form is not a younger version of the base referent. Another example of a similative diminutive is puoskǝn ‘swaddle’ in (27), which is derived from puos ‘bag; vulva’. We assume that the meaning ‘swaddle’ is derived from the meaning ‘bag’ because of the similarity in shape and function.
Note that our category similative is slightly broader than Jurafsky’s (Reference Jurafsky1996) category ‘imitation’, which covers cases where the diminutive form refers to an imitation or copy of the base referent. While some examples in our material can be considered cases of imitation, e.g. bååtjǝn ‘bib’ from båårt ‘beard’, we prefer to analyse these as instances of a broader similative category.
The two constructions are also both found in meronymic contexts where the diminutive noun refers to a part of a whole, though this is rare in the material. The -kǝn diminutive in (28) is the only example of this construction in meronymic use. The base word rind refers to the entire crust of a loaf of bread, while the diminutive rintjən refers to the heel, i.e. a part of the crust:
With the gender-shift diminutive there are two examples of meronymic use, given in (29). With common gender, muos refers to a bog in its entirety, while neuter muos means ‘peat’. We interpret this as a meronymic relation, peat being one of the constituent parts of a bog. Similarly, common spoan refers to an entire layer of wood, while the neuter compound hevlspoan refers to the smaller fragments of this that are shaved off with a plane.
Thus far, we have discussed functions that are attested for both diminutive constructions. We now turn to the two functions that we only found with the gender-shift diminutive and the one function which we only found with the -kǝn diminutive. We begin with a function – the singulative one – which is only securely attested once; see (30).
Common sieb is a mass noun referring to soap as a product, whereas neuter sieb is a count noun referring to an individual piece of soap. Bjerrum & Bjerrum (Reference Bjerrum and Bjerrum1974, s.v. sæbe) record a few minimal pairs like (31) to illustrate the difference. Note common dænd and guo in (31a) vs. neuter gåt in (31b).
The other function which is unique to the gender-shift diminutive in the material is the pejorative one, i.e. negative speaker attitude. This was noted by Bjerrum & Bjerrum (Reference Bjerrum and Bjerrum1974, s.v. en), though they do not mention that this meaning is not recorded for the -kǝn diminutive. The following are examples where the gender shift is explicitly said to have pejorative meaning:
The noun kvind with common gender is the neutral term for an unmarried woman, but with neuter gender it takes on a disapproving meaning and implies that the woman is somehow irresponsible or improper. In the example in (33a) it refers to a young woman who is disorderly, in (33b) to someone who is promiscuous.
With some other lexemes, it is not explicitly stated that the neuter form was pejorative, but the available examples suggest that this was its function. Compare kvindmindǝsk, another word for ‘woman’ (lit. ‘woman person’), which has common gender in the neutral context in (34a), but neuter gender in the disapproving expression in (34b).Footnote 13
This use of diminutive constructions to refer disparagingly to women has been observed in many other languages (Jurafsky Reference Jurafsky1996:547–548). However, note that in Viöl Danish, pejorative diminutives do not appear to have been limited to female referents; the compound kaarlmindǝsk ‘man’ (lit. ‘man person’) is also recorded with neuter gender, like kvindmindǝsk in (34b), and nusǝlatsk ‘dawdler’ (see [32]) is not reported to be restricted to women.
The only function which is unique to the -kǝn diminutive is the hypocoristic one, i.e. a type of positive speaker evaluation. The -kǝn diminutive of mand ‘man’ is used as a term of endearment for a boy, and from huund ‘dog’ one can derive the affectionate diminutive huntjǝn (also mentioned in [19] above).
In the following nursery rhyme, one of the only attested examples of child-directed speech, there are several diminutives whose function is clearly mainly affectionate; the rhyme is accompanied by gestures towards the child’s chin (tjabtjǝn, see [21] above), mouth, cheek, nose, and eye, after which it nods its head.Footnote 14
The -kǝn diminutive may also have affectionate connotations even when it occurs in one of the other functions. The form hierǝkǝr in (37), from a short verse, is recorded as referring to shepherd boys (i.e. a younger referent), but it clearly occurs in an affectionate context.
Note, finally, that the -kǝn diminutive was very productive in the formation of nicknames (Bjerrum Reference Bjerrum1934, Reference Bjerrum1944:222–226), which we assume were generally hypocoristic. Two examples from the texts are matǝskǝn (diminutive of Matthes ‘Matthew’) and hanskǝn (diminutive of Hans ‘John’) (Bjerrum Reference Bjerrum1975:84).
The semantic analysis has shown that there was a great deal of functional overlap between the two diminutive constructions. They are both attested with the meanings smaller referent, younger referent, classificatory, similative, and meronymic. The gender-shift diminutive is also recorded with the meaning singulative, but only once, so we cannot say if this was a real difference between the two constructions. However, a clear difference emerges within the field of qualitative meanings, i.e. speaker attitude: whenever one of the two constructions expresses speaker attitude, the gender-shift diminutive always has pejorative meaning, and the -kǝn diminutive always has hypocoristic meaning. Figure 1 schematizes the functions that are recorded for the two constructions in the data.
Attested meanings of gender-shift (—) and -kǝn (- - -) diminutive.
There are only a few words in the material which are found with both constructions. The word huund ‘dog’ is worth citing again because it neatly illustrates the difference in speaker attitude; compare the meanings in (38).
The gender-shift diminutive huund is reported to have pejorative meaning, as in (39a), whereas the -kǝn form huntjǝn is attested in several quotes where the main consultant is speaking affectionately about or to her dog, as in (39b).
It should be stressed that the two constructions do not always have qualitative connotations. For some other lexemes where both diminutives are recorded, no functional difference between them is mentioned, or the difference does not appear to be related to diminution. Compare the two diminutive formations from myø; the gender-shift form is reported to refer to a school-aged girl, whereas the -kǝn form refers to an even younger girl:
In this case, there is no indication in the material that the gender-shift diminutive had negative connotations. The following quotation is from one of the consultants talking about herself:
Thus, the difference between the constructions was not that gender shift was always pejorative and the -kǝn form always hypocoristic. Rather, the difference lies in their functional potential: gender shift could be used to express pejorative meaning, and the -kǝn diminutive could be hypocoristic – but apparently not the other way around.
5. The role of language contact
We have argued that Viöl Danish had a diminutive system with two available morphological constructions, a situation which appears to be unique among the Scandinavian languages. We will now briefly consider how this system may have arisen, in particular how contact with Low German may have shaped it.
As for the suffix -kǝn, there is general agreement in the literature that it must have been borrowed from Low German, the main contact language of Viöl Danish (Bjerrum Reference Bjerrum1931:124, Nielsen Reference Nielsen1959:63, Hofmann Reference Hofmann1961:73).Footnote 15 This may at first appear somewhat surprising, as the northernmost dialects of Low German have frequently been cited as making no or only limited use of diminutives (Wrede Reference Wrede1908:91, Mensing Reference Mensing1927–35, s.v. -kǝn, Keller Reference Keller1961:342). As Hofmann (Reference Hofmann1961:61–68) shows, however, this is a relatively new state of affairs, as diminutives are amply attested in northern Low German textual sources from the seventeenth and eighteenth centuries and in the dialect dictionary by Schütze (Reference Schütze1800–06). Of course, the suffix must have entered Viöl Danish before it became obsolete in the Low German contact dialects, and there are indeed good reasons to assume that it was not a recent innovation in Viöl Danish. First, as we noted above, there are both phonologically conditioned allomorphs (-jǝn and -tjǝn) and several irregular diminutive forms (e.g. möjkǝn ← myø); this is unexpected if the form had been borrowed within the last few generations. Second, there are -kǝn forms with lexicalized meanings which are not immediately predictable from the base word (e.g. puoskǝn ‘swaddle’ ← puos ‘bag; vulva’); this semantic diversity must also have had some time to develop. Third, -kǝn was clearly widely adopted in the community to form nicknames and place names (see note 8), and Bjerrum (Reference Bjerrum1934) found several names with -kǝn in a list of farm names in Viöl parish from 1798 (e.g. Lenckens, genitive diminutive of Lene, and Hanschen, diminutive of Hans). We can thus assume that the formation was already established in Viöl Danish at least by the end of the eighteenth century.
Another question is how the suffix was borrowed into Viöl Danish. It is generally assumed that lexical borrowing is ‘a pre-requisite and a trigger for morphological borrowing’ (Matras Reference Matras2009:211). In other words, borrowed affixes do not enter a language on their own, but as part of loanwords, from which they may become productive and start appearing on inherited words, so-called ‘backwards diffusion’ (Matras Reference Matras2009:209). We indeed find many Low German loanwords among the -kǝn forms in the material. These include both apparent diminutiva tantum with no corresponding base word (e.g. kaninkən ‘rabbit’, laṅtgən ‘plough wheel’, løpəkən ‘marble’; see also note 10) and pairs of base word and diminutive (e.g. båårt/bååtjen ‘beard/bib’, botǝl/botǝlgǝn ‘bottle/small bottle’, jong/jonkən ‘boy/little boy’). The number of such loanwords must have been sufficiently large for the formative -kǝn to be recognized and adapted as a productive affix within Viöl Danish itself.
The development of the gender-shift diminutive, on the other hand, does not appear to be a direct result of language contact. While the use of gender shifts to express evaluative meanings is cross-linguistically widespread (Di Garbo Reference Di Garbo2014:145–176, Vanhove & Ahmed Reference Vanhove and Hamid Ahmed2018:54–55, Steriopolo Reference Steriopolo2021), we are not aware of any comparable diminutive constructions in the immediate neighbours of Viöl Danish. In this case, we suggest that the diminutive was an internal development motivated by the almost exclusive use of the neuter when referring to children and animal offspring in the dialect. Not only are most nouns referring to young individuals neuter, the neuter was also used when referring anaphorically to children and young animals, as in the examples in (42).
Because of this association between the neuter and the semantic feature ‘young’, neuter gender assignment became a strategy for referring to younger referents. From there, it was gradually extended to referents of smaller size and to the other meanings discussed above. Whether this happened before, after, or concurrently with the development of the -kǝn diminutive is impossible to say. It is quite likely, however, that language contact had also played an indirect role in the development of the gender-shift diminutive. The association between neuter gender and human and animal offspring appears to be an areal feature which the dialects of southern Denmark share with Low German and North Frisian (see Löfstedt Reference Löfstedt1968:9–10). A detailed comparison of the gender systems of these languages is beyond the scope of this study, but our conjecture would be that the southern Danish and North Frisian systems were, at least in part, shaped by Low German influence. In this way, the gender-shift diminutive would also ultimately have its origins in language contact, although of course in a much less direct way than the -kǝn diminutive.
6. Conclusion
We hope to have shown that Viöl Danish – unlike the modern Scandinavian languages – had a productive diminutive system which was still used by the last speakers in the early twentieth century. We have argued that the dialect had two distinct diminutive formations, one involving a suffix and the other a gender shift from common to neuter. We found that there was a great deal of functional overlap between them – both are attested with roughly the same range of meanings – but a clear difference emerged within the domain of qualitative evaluation: while the -kǝn diminutive could be used to express affection, including in nicknames, the gender-shift diminutive could be used with pejorative meaning. Finally, we suggested that language contact played a direct or indirect role in the development of both constructions: -kǝn is a clear case of borrowed morphology from earlier Low German, while the gender-shift diminutive was motivated by the same association between neuter and youth which is also found in Low German and North Frisian.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S0332586526100754
Acknowledgements
We are grateful to Olga Fischer, Jose A. Jódar-Sánchez, Christoph Winter, and three anonymous reviewers for their valuable feedback and encouragement. Sune Gregersen’s work was made possible by a grant from the Ulla and Børge Andersen Foundation for Linguistic Research.
Competing interests
The authors declare none.
Open access
