CrossRef citations to date
Sémantique lexicale/ Lexical semantics

Endocentrism and analytic naming as communication strategies of the Danish language

Pages 127-146 | Received 31 Oct 2022, Accepted 28 Sep 2023, Published online: 03 Jan 2024


This article proposes a correlation between Michael Herslund’s concepts of “endocentric and exocentric languages” and Per Durst-Andersen’s concepts of “hearer-oriented and reality-oriented languages”, correspondingly. The article argues that the Danish syntagmatic tendency to discrete expression of meanings is manifested both in verbal and nominal naming. This syntagmatic tendency is interpreted as a manifestation of an addressee-friendly naming strategy, since it allows the addressee to reconstruct the meaning by referring to a general semantic category. The sections of the article are devoted to the implications of analytic naming as a reflection of the Danish hearer-oriented communication strategy and syntagmatic tendency to discreteness, modularity and distancing in expressing complex meanings. In this regard, the following phenomena are considered: systemic analytical tendencies in the Germanic languages, manifested at different language levels; analytical tendencies in the structural and semantic organization of Danish compounds; the modular nature of Danish verbal compounds and their correlation with analytic complexes.

1. Introduction

At the turn of the 20th and 21st centuries, a unique community of research linguists formed at the Copenhagen Business School (CBS), bringing together leading experts in the field of Romance languages, Russian, English, as well as the theory and practice of translation. In 2001, the author of this article was granted an internship in Copenhagen and got acquainted with colleagues from the Department of French, Italian and Russian languages, as this section was then called. The think tanks were Professor Michael Herslund, Head of the Romanсe languages team, and Professor Per Durst-Andersen, Head of the Russian team. Almost simultaneously, they have been developing their typological theories: Michael Herslund worked on the theory of endocentric and exocentric languages (Herslund Citation1994, Citation2000, Citation2002) and Per Durst-Andersen developed the theory of communicative typology of languages (Durst-Andersen Citation1995, Citation1997, Citation2011). At first glance, these theories seemed to have nothing in common, but I shall argue that, in fact, they organically complemented each other.

In this article, I will show how the concepts of “endocentric and exocentric languages” put forward by Michael Herslund correlate with the “hearer-oriented and reality-oriented languages” theory suggested by Per Durst-Andersen.

Michael Herslund, in collaboration with Irène Baron, made an accurate observation that in all modern Romance languages concrete lexical nomination is provided by nouns, while verbs present much more abstract meanings. This becomes especially evident, for instance, when comparing translations from Romance languages into Danish and vice versa (Baron Citation2002; Baron and Herslund Citation2005; Herslund Citation2000, Citation2002; Herslund and Baron Citation2003).

A compendium, based on Michael Herslund’s theory (Korzen and Lundquist Citation2005) was devoted to this topic. This book shows that:

De germanske sprog bruger ret præcise verber mens deres substantiver i forhold til de romanske fremtræder som underspecificerede og upræcise. Man kan derfor sige at de præcise informationer i den germanske sætning findes i sætningens centrum, verbalet, og man kan derfor kalde disse sprog endocentriske (endo ‘inden i’ + ‘centrum’). De romanske sprog derimod betjener sig af abstrakte verber og præcise substantiver. Den mest præcise information lokaliseres i substantiverne, ikke i sætningens centrum, verbalet, hvorfor vi kan kalde dem exocentriske sprog (exo ‘uden for’ + ‘centrum’). (Korzen and Lundquist Citation2005, 19)

– The Germanic languages use fairly precise verbs, while compared to the Romance languages, their nouns appear underspecified and imprecise. You can therefore say that the precise information in the Germanic sentence is found in the center of the sentence, the verbal, and you can therefore call these languages endocentric (endo ‘within’ + ‘centre’). The Romance languages, on the other hand, use abstract verbs and precise nouns. The most precise information is located in the nouns, not in the center of the sentence, the verbal, which is why we can call them exocentric languages (exo ‘outside’ + ‘centre’).

In this article, I propose to approach the distinction between endo- and exocentrism not in the lexico-semantic, but in the communicative-pragmatic aspect. Pragmatics is understood as a section of linguistics that focuses on the study of the relationship between linguistic signs and their users in the process of communication.

2. Reality-oriented vs. hearer-oriented naming strategies in exo- and endocentric languages

The world around us, the denotative reality, is prototypically objective. Languages tend to describe reality as accurately as possible, ideally, providing a separate word for each object of the surrounding world. The Romance languages strive for maximally succinct and accurate naming. As a result, the comparison shows that many French lexemes, for example, pichet, cruche, pot, broc correspond to just one Danish lexeme kande. As Baron and Herslund point out, the Danish word kande is an “underspecified noun < … > that covers a large number of entities with a limited number of features in common (the Danish word kande, for example, can be used for any container with a spout for pouring liquid from)”, while in French it only corresponds to “specified nouns”, which “denote a more limited category of objects with many common features” (Baron and Herslund Citation2005, 29).

However, this does not mean that the Danish language does not differentiate between these objects in reality. Danish offers a wide range of compound names to distinguish different functional purposes of the respective objects (vinkande, vandkande, mælkekande, vaskekande). In order to understand French utterances containing the words pichet, cruche, pot, broc, the addressee must know them in advance, that is, imagine how the corresponding objects look in reality. In order to understand Danish utterances containing the words vinkande ‘wine jug’, vandkande ‘water jug’, mælkekande ‘milk jug’, and vaskekande ‘wash pitcher’, the addressee does not need to have any advanced knowledge. Knowing the non-specific meaning of the word kande ‘jar’, the hearer can reconstruct the function, for example, the meaning of the word sovsekande ‘sauce pot’ by analogy with mælkekande ‘milk jug’. That is, in its naming strategy, Danish is more kundevenlig ‘customer friendly’, or more precisely addressee-friendly than French. This fully corresponds to the communicative typology of the Danish language as a hearer-oriented language.

In the verbal system of the Danish language, the orientation towards the recipient of the utterance is even more intense. Danish is quite endocentric, in other words – verbocentric. But in modern Danish, predicates are increasingly and consistently expressed by discrete units analytical combinations of lexically charged verbal lexemes with auxiliary verbs, adverbs of direction and modal particles that specify the relationship between communicants and the proposition. The two examples below show the difference between the two modal particles nok (expressing the speaker’s conclusion) and vist (expressing external evidentiality).

The point is that a wide range of different analytical means, widely represented in Danish, serve the purpose of specifying subjective orientation – in space (spatial adverbs – as in svømmet væk), in time (analytical verbal forms as har svømmet vs. er svømmet) and in communication (modal particles as nok and vist). Research on Danish modal particles, realized by Elvira Krylova (Citation2021), proved that Danish modal particles differ systemically, referring epistemic responsibility either to the speaker (nok), or to the speaker+ the hearer (vel), or to external evidentiality (vist).

A comparison of Danish and French, carried out by Baron and Herslund (Citation2005) showed that the French verb can be more abstract than the corresponding isolated verbal Danish lexeme. Compare French: le chien entra, le poisson entra, l’oiseau entra, la voiture entra, le navire entra and Danish: hunden gik/løb ind, fisken svømmede ind, fuglen fløj ind, bilen kørte ind, skibet sejlede ind (27). But if we look at Danish predicates as analytic words (in the examples above – as phrasal verbs), then the pattern is exactly the same as the pattern noted above regarding Danish nominal compounds. The final component of Danish phrasal verbs defines an abstract category, in this case “inward movement”, while the initial component specifies the event, makes it concrete, just as the first component of the compounds mælkekande and sovsekande specifies the concept of kande. This is exactly the same “addressee-friendly” way of naming, because even if the recipient, for example, is unfamiliar with the verb humpe, he or she will still understand that the sentence Læreren humpede ind i klassen ‘The teacher hobbled into the class’ describes the direction of the subject’s movement into the locus.

2.1. Addressee-oriented movement description in Danish vs. reality-oriented movement description in Russian

While, as shown above, French verbs can be more abstract than the corresponding isolated verbal Danish lexemes, Danish verbs typically appear to be more abstract than the corresponding Russian verbal lexemes. In the case of indirect equivalents, the discrepancies between Russian and Danish can always be seen as a more abstract description of movement in Danish and a more detailed and nuanced description of movement in Russian. This regularity was noted by Viktor Smith in the chapter “Tilfældet russisk”/“The Case of Russian” (Smith Citation2005, 202–206).

To prove this, it is enough to compare original Danish texts with their literary translations into Russian made by professional translators. Let us turn to the translation into Russian of Ida Jessen’s (Citation2009) novel “ABC” (Citation2005):

The systemic comparison of parallel texts shows that Russian and Danish languages focus on different ideas. In Danish, the orientation of movement against the observer is more important than the quality of the movement itself. In Russian, the situation is quite opposite. Special lexemes help to depict a specific character of a certain movement.

The Danish original uses verbs that indicate only the generalized type of action: hænge – “fasten an object on or around something that is raised above the ground or floor, e.g., a rope, hook, pole or body part” (DDO Citation2022); komme – “move or travel to a specific place or to the place where the speaker is” (DDO Citation2022). If we focus on the Russian equivalents, it turns out that the Russian translation semantically specifies the nature of the subject’s movements in relation to the objects of the external world. Russian thus behaves like a typical reality-oriented language, focused on reflecting denotative reality, and strives to convey the mode and nature of the movement as accurately as possible.

In the first example, the verb razvesil (PST.3SG.) clarifies the distributive nature of the subject’s movements when objects are placed vertically relative to the surface. The second verb, pot’anutsa (FUT.3PL), specifies the unhurried nature of a dense car traffic, when cars “move slowly one after another, in a row” (BTSRYA Citation1998).

However, the comparison makes it clear that in the Russian translation the subject-orientational component of the original Danish meaning is weakened. This is due to the fact that the Danish language regularly uses phrasal verbs that categorize the direction of action through analytical nomination. Thus, in the first example, the movement denoted by the first predicate (hængte ud over rækværket) is directed outward (ud) from the subjective space of the person (in this case, the space of the house), and the second predicate (kiggede ned i haven) specifies the vertical orientation of the object relative to the person’s point of view. The addressee can see clearly that the person looks at the garden from the top (ned), that is, apparently, from the top floor of his house. The second example (ville komme forbi), specifies the trajectory of movement towards the person (komme) and the further direction of movement (forbi) beyond the point where that person is located.

This means that apart from the spatial orientation of verbal actions relative to the external objects, which is always preserved in Russian translations (cf. Dan. over rækværket – Rus. na perilakh LOC.PL.; Dan. i haven Rus. v sad ACC.SG.) the Danish sentence regularly mentions the orientation of the action relative to the subject. And this very meaning is omitted in the Russian translation. By “subjective orientation”, I mean not only the spatial orientation relative to the speaker’s location (the 1SG in direct speech), but also orientation relative to the position of a character in the narrative text.

An indication of the spatial position of objects relative to the subject gives the addressee the opportunity to literally “put himself in place” of the subject, to see and feel how the character’s space is organized in the narrative (or the speaker’s space in a situation of direct verbal communication). Thus, by means of the analytical nomination of the Danish language, a communicative effect is created, which I propose to call “spatial empathy”.

The fact that the Danish predicate conveys additional information about the spatial orientation of the verbal action fully corresponds to the communicative type of the Danish language designated by Per Durst-Andersen as a “hearer-oriented language”.

3. Analytical tendencies in the Germanic languages manifested at different levels

The linguists have long noticed the generalized, unidirectional tendency of the Germanic languages, that proceeds with varying degrees of intensity, towards root isolation, agglutination and analyticism (Guchman Citation1981; Plotkin Citation1989).

Initially, these trends were identified in the dynamics of morphological systems; however, since the 90s the focus of attention has shifted to the presence of these diachronic constants at other levels of the language system. For instance, Kuzmenko (Citation1991) clearly showed how these trends manifested themselves in the evolution of the phonological system of the Germanic languages. The dissertation by Shaposhnikova (Citation1999), based on material of the English language, focused on interlevel lexical-syntactic processes and related phenomena of analytical naming. It was shown that such parameters as binary structure, productivity of modeling, and the preservation of semantic transparency in each of the constituents, introduce analytical verb binomials (structures like verb–verb start singing, verb–adjective get rich, verb–noun have a cry, verb–adverb carry out) into the sphere of systemic verbal lexeme formation.

In Danish, examples like the English analytical verbal binomials above are also plentiful. In Russian, they regularly correspond to one-word equivalents that name a specific action. Compare, for example, the structures Vfin+Vinf like Dan. få at vide ‘get to know’ and Rus. uznavat’; V+Adj like Dan. gøre rent ‘to clean’ and Rus. ubirat’sya; V+N like Dan. tage fejl ‘be wrong’ and Rus. oshibat’sya; V+Adv like Dan. sige op ‘to quit’ and Rus. uvoln’atsya.

This article aims to demonstrate that the syntagmatic tendency towards the analytical expression of meanings is a manifestation of the addressee-friendly strategy, since it allows the addressee to reconstruct the meaning by linking it to one or another general semantic category. Below, I will show that analytic tendencies manifest themselves in Danish not only at the level of analytic phrases, but also at the level of compound words, expressed in structural and semantic discreteness of their components.

It may seem that one can speak about analytical trends in compound word formation only in relation to the English language. In English, typical complex words are “unstable compounds”, while in the Scandinavian languages, the compound word formation still retains unshakable positions.

The term “sochetaniya slov nestoykogo tipa” (“unstable compounds”) was introduced by Smirnitsky (Citation1956) referring to N+N structures as stone wall. Other linguists consider such examples to be “free phrases” but still notice that: “The linguistic analysis of extensive language data proves that there exists a regular correlation between the system of free phrases and all types of subordinative (and additive) compounds” (Ginzburg et al. Citation1979, 151).

Cf. composition of the same words from English (COD Citation1995) and Danish (NDO Citation1996) dictionaries:

  • English merged compound snowflake = Dan. snefnug (cf. Rus. one-stem word snezhinka); Eng. snowman = Dan. snemand (cf. Rus. one-stem snegovik);

  • Eng. hyphenated. snow-drift vs. Dan. merged snedrive (cf. Rus. one-stem sugrob);

  • Eng. unstable compound snow owl vs. Dan. merged sneugle.

Nevertheless, even for merged Danish compound words, there exists an analytical tendency to strengthen the discrete character of the unit. It is not just about connecting word stems, which is also natural for inflectional languages, but about a rising trend towards discrete elements in the composition of the Danish compound word.

3.1. Discreteness in the structural organization of Danish compounds

At the structural level, the relative independence of parts in the Danish compound is manifested, in particular, in the stress on each of the parts and in the absence of fusion phenomena at the root junctions. The modifying element of the Danish compound, as a rule, is presented as an independent unit. Cf. Russian compounds with either a connecting vowel at the junction of stems (snegoochishcháyushij ‘snow-clearing’, svezhevýzhatyj ‘freshly squeezed’), or assimilative phenomena at the root junctions – in less typical abbreviated compounds, for example, regressive assimilation of voiced consonant in hozblók [-zb-] ‘household shed’, and deaf consonant in hoztováry [-st-] ‘household goods’. Moreover, a unifying stress on the last stem is typical for Russian compounds.

In this regard, it is interesting to compare the Danish language with the Swedish, which has made relatively less progress along the path to analyticism. In contrast to Danish, the reduction in Swedish proceeded less radically, and the relics of old nominal stems were sometimes preserved. These are the elements -a-, -u-, -o-, usually described as connecting vowels in Swedish compounds. However, they differ significantly from such productive connecting elements in Danish as -s- and -е- in that they retain the historical form of the noun, which in modern Swedish does not occur in isolated words.

This structure indicates the lack of morphological independence of the component, since in modern Swedish such obsolete forms can only be found as part of complex words: fiska|fange ‘fishing’ < fisk ‘fish’ (preserving vowel -a- of Germanic Masc. -a- stem, cf. obsolete Swed. Pl. Nom. fiskar); kyrko|råd ‘church council’ < kyrka ‘church’ (preserving long vowel -o- from Germanic Fem. -o:n- stem, cf. Pl. kyrkor).

The corresponding Danish words retain only a pure stem that is equivalent to the original form of a separate word. Cf. Swed. kvinno|tycke < kvinna and Dan. dame|tække < dame ‘ability to attract women’. (Danish-Swedish parallels are given according to Norsteds Skandinaviska Ordbok (NSO Citation1994)).

The fact that the modifying component within the compound is perceived as more discrete, while its analogue may be used as a separate word, corresponds to the analytical tendencies that are more active in Danish.

The tendency to correlate the component of the compound with the original form of the word in its independent usage is especially noticeable in compounds where the first element is verbal. Here, we can talk about the tendency in the Danish language to replace stem-composition with word-composition.

In this connection, it should be clarified that in modern Danish, the reduction of the common Scandinavian ending of the infinitive -a into -e created the homonymy of the old connecting vowel -e- and the infinitive ending -e. This “side effect” of the phonetic process is functionally valid due to the fact that the first element of such compounds retains a semantic connection with the verb, for example, Dan. føle|horn ‘snail horns’ <horns to feel> or kravle|gård ‘playpen’ <courtyard to crawl>. So, in modern Danish language, the first (verbal) part of the compound completely coincides in its form with the infinitive.

It is possible to prove that verbal compounds retain traces of the infinitive and not of an undifferentiated root with a connecting vowel e by analyzing examples where the verbal and nominal stems differ formally in the sense that they have alternating vowels in the root. I conducted the following experiment: from the vocabulary list of verbs with root vowel alternations, all verbs were selected that correspond to nouns with a different root vowel. From NDO, compounds formed from the corresponding substantive and verbal stems were selected. All in all, eight correlations of noun/verb stems were examined. A total of 212 related compounds were found in (DDO Citation2022): brud- 3, bryde- 2; dom- 5, dømme- 3; frost- 16, fryse- 12; gråd- 4, græde- 5; røg- 28, ryge- 18; sang- 34, synge- 5; skud- 20, skyde- 29; tal- 26, and tælle- 2. The result turned out to be very significant: absolutely all stems that were clearly identifiable as substantive were included in the compound as an isolated noun (without a connecting vowel), and all unambiguously verbal stems were included in the compound in the form of an infinitive, that is, they ended in -e. There was not a single case recorded where the verbal compound was formed from a pure root and not from the infinitive. Here are some examples:

The above list confirms our hypothesis that in modern Danish the verb is included in the compounds as the first element in an isolated form of the infinitive.

The author of this article observed a curious evidence of how native Danish speakers perceive the transparency of the meaning of a separate word in a compound in a conversation between a foreigner and a Dane in Copenhagen. The foreigner noticed cars with ice cream advertisements “Hjemis” and assumed that this ice cream was home-made (from hjem “home”). His Danish companion corrected him: “No, the name of the company is Hjemis and not Hjemmeis This is ice cream that is delivered home (leveret hjem), not made at home (lavet hjemme)”.

The root of the word is the same in the noun hjem “home”, and in the adverb of direction hjem “home”, and in the adverb of location hjemme “home”. But the meaning of the compound is not made up of the meanings of roots, but of meanings of constituent stems that match separate words. This serves as a clear illustration of the addressee-friendly structuring of Danish compounds, when the very structure of the compound allows the addressee to more accurately decode its meaning.

Thus, though Germanic analytical tendencies are common to all Scandinavian languages, contrastive analysis reveals a greater discreteness (syntacticity) of components in the Danish language compared to higher contact cohesion (morphologization) of compound words in Swedish. This confirms a particularly high level of structural discreteness of Danish compounds.

3.2. Modularity of Danish verbal compounds: correlation with analytic units

Michael Herslund noted that accurate information in Germanic languages, compared to Romance languages, is concentrated in the verb, that is, in the center of the sentence, for which reason these languages are called “endocentric”. It is important for us to show that the elaboration of lexical information (as noted above) in the Danish verb often follows the path of nominal compounds. In verb compounds, just as in nominal compounds, the first component specifies the lexical meaning and the second component (abstract verb like gøre ‘to do’, sætte ‘to put’, give ‘to give’, lægge ‘to put’) categorizes grammatical meanings, for example, causation. The recipient can easily decode their meaning due to their modular structure and predictable correlation with phrases: gøre fri frigøre ‘to free’ Rus. osvobozhdat’; gøre vanskeligt vanskeliggøre ‘to complicate’ Rus. zatrudnyat’; sætte i pant pantsætte ‘to pawn’ Rus. zakladyvat’; sætte i stand – istandsætte ‘to repair’ Rus. chinit’; give navn navngive ‘to name’ Rus. imenovat’; give form formgive ‘to form’ Rus. formirovat’; give til kende – tilkendegive ‘to notify’ Rus. uvedomlyat’; lægge til rette – tilrettelægge ‘to fix’ Rus. nalazhivat’; lægge bro brolægge ‘to pave’ Rus. mostit’; lægge grund til – grundlægge ‘to establish’ Rus. osnovyvat’; lægge planer – planlægge ‘to plan’ – Rus. planirovat’. I give Russian correspondences everywhere, showing that these complex meanings are regularly conveyed in Russian by a single stem lexeme, with which the recipient must be familiar in advance, while the Danish recipient can reconstruct the meaning of the predicate from the meaning of its constituents.

This situation is reminiscent of the agglutinative word, which “is not stored in memory in a finished form, but every time is “constructed” anew from “brick”-morphemes. As a rule, the root morpheme has a high degree of independence and can function as a separate word” (Ilyina Citation2004, 55). I propose to use the term “modularity” for the correlation between the components in a compound and a separate word that is present in the mind of a speaker. Discrete modules can form analytical structures, or can be combined into complex units (Dan. sammensætninger).

Modularity also characterizes such an important area of Danish verbal word formation as phrasal verbs. Danish clearly distinguishes prefixes that do not correspond to separate words (for example, undgå ‘avoid’, forbedre ‘improve’, erklære ‘declare’), and preverbs that correspond to prepositions and adverbs. These are examples like binde om - ombinde ‘to bind’, høre til - tilhøre ‘to belong’, kalde ind - indkalde ‘to convene’, skrue af - afskrue ‘to unscrew’, tage til - tiltage ‘to strengthen’. It is significant that in the Danish linguistic tradition both preverbial and postverbial variants are defined as “compound verbs” (sammensatte verber), which differ in the location of the modifying component (synthetic or analytical) and are called, respectively, “true and untrue verbal compounds” (ægte og uægte verbale sammensætninger) (Diderichsen Citation1976, 236–239). This confirms that the idea of potential splitting, analyticity of Danish synthetic forms is present in the linguistic consciousness of native speakers.

Even when the meaning of the synthetic form is changed from a more concrete to a more abstract one (komme ud ‘to go out’ and udkomme ‘to be published’, føre ud ‘to bring out’ and udføre ‘to export’) up to the splitting of polysemy (komme ned ‘to go down’, but nedkomme ‘to give birth’, gå under ‘to sink’, ‘to perish’, but undergå ‘to endure’), the correlation with the adverb remains unchanged. And even in the absence of a parallel construction (for example, falde ned ‘to fall down’ versus *nedfalde), the analogy with the productive model of the correlation (like indstille stille ind ‘adjust’, indgå ind ‘enter’) is still present.

The discreteness of the preverb as part of a complex word is also supported by the stress. Unlike typical prefixes, the adverbial component of a verb retains a strong stress in both pre- and post-verbal usage, cf. er’holde ‘get’, ‘achieve’, but opholde ‘suspend’, ‘delay’ and holde op ‘stop’.

3.3. Analytical tendencies in the semantic organization of Danish compounds

Thanks to the productivity of compound word formation, it is impossible to even approximately count the number of words that exist in modern Danish, since a huge number of complex words are constructed from “modules” right “on the spot”. And although such newly-constructed words are perceived by native speakers as correct, they have very little chance of being included in the vocabulary.

Such compound words are usually called potential. To illustrate the activity of “potential word formation” in Danish, I selected a random text (an international news column from the newspaper “Berlingske tidende” in 50 lines) and compared the ratio of simple and complex (vocabulary and potential) words.

As a result, out of 155 unrepeated simple words (nouns, verbs, adjectives, adverbs), 18 were compounds registered in the dictionary: baggrund ‘background’ (n), bosætte ‘settle’ (v), domstol ‘court’ (n), forfatningsstridig ‘unconstitutional’ (adj), generalsekretær ‘general secretary’ (n), igangværende ‘current’ (adj), krigsforbryder ‘war criminal’ (n), krigsofre ‘war victims’ (n), millionbeløb “million amount” (n), naboland “neighboring country” (n), nødhjælp ‘humanitarian aid’ (n), offentliggøre ‘make public’ (v), premierminister ‘prime minister’ (n), pressemøde ‘press conference’ (n), sagsøger ‘plaintiff’ (n), søgsmål ‘claim’ (n), tocifret ‘two-digit’ (adj), verdenskrig ‘world war’ (n). Other eight complex words in the text were not registered in the dictionary and thus can be categorized as potential words. These are either two-term compounds røveri|udbytte “proceeds from loot”, udleverings|aftale ‘extradition agreement’), or polynomials (distrikts |dom| stol ‘district court’, menneske |rettigheds| eksperter ‘human rights experts’, nød |hjælps| arbejdere ‘employees of humanitarian aid organizations’), or hyphenated with proper names (Dafur-region ‘Dafur region’, FN-folk ‘UN staff’, Yasukuni-tempel ‘Yasukuni shrine’). To be productive potential words must be semantically transparent and their meaning must be easily deduced from the meaning of their components. And this, as mentioned above, is indicative of the analytical naming and an addressee-oriented language.

An art installation that the author of this article once happened to see at Louisiana Museum of Modern Art serves as a curious example of transparent semantic motivations and productivity of Danish composite neologisms. Each visitor could come up and take a multi-colored small stone from one of the baskets exhibited by the artist. Each basket had a wish inscription and a name that the artist came up with for each type of stones. For instance, one inscription read: Huskesten. Med den kan vi huske alt det vi vil huske. “Memory stone” (lit. “remember-stone”) “With it you can remember everything you want to remember”. And the message on another basket was: Ledesten. Den vil lede ham som ejer den til de rigtige steder. “Guiding stone” (lit. “lead-stone”) “It will lead the owner to the right place”. It is clear that the names of stones were occasional neologisms created by the author, constructed around an infinitive naming and attributed to each type of stones. There was no doubt that each neologism would be properly understood, since the word-building model ensured absolute semantic transparency.

Modularity, syntacticity of stringing meanings in Danish compounds is clearly visible from the following example given by Erik Hansen: “Let’s say you find that the word oplysninger ‘information’ has a not very specific meaning in this context, then you clarify and write bogholderi|oplysninger ‘accounting information’. If this is not enough, then you can add something else, for example løn|bogholderi|oplysninger ‘salary accounting information’. It is quite possible to continue in this spirit: ferie|løn|bogholderi|oplysninger‘accounting information on vacation pay’, sommer|ferie|løn|bogholderi|oplysninger ‘accounting information on salary for summer vacation’ (Hansen Citation1996, 63, my translation).

It is important to emphasize that semantic transparency is an essential prerequisite of productive models. In modern Danish, there are very few old semantically fused compounds like solsort ‘starling’ (literally ‘sun-black’), but the trend is set precisely by productive models. A clear pattern can be identified: the more productive the model, the more evident is the functional differentiation of its components, when, the last component develops a generalizing or categorizing function. And this, as noted at the beginning of the article, is how analytical naming is manifested.

The main direction of functional differentiation in compound word formation is the categorization of hypo-/hyperonymy between the key word and various compounds derived from it.

By comparing the principles of nomination in Danish and Russian, we can uncover the interlingual differences in the nomination and match the Russian language with the “reality-oriented”, and the Danish language with the “hearer-oriented” types. The Danish-French and Danish-Russian comparison reveals similar word-formation strategies: in case of discrepancies, Russian and French choose the integrative naming of a new denotation by a new word, while in Danish the most productive is the compound naming of a new notion through categorization by means of a broader concept.

Let us illustrate this thesis by comparing the naming of shoes in Russian and Danish. In Russian, the speaker identifies qualitatively different objects by using special simplex words, for example: valenki ‘felt boots’, krossovki ‘sneakers’, tapki ‘slippers’, cheshki ‘dancing shoes’, sandalii ‘sandals’, puanty ‘pointed shoes’, platformy ‘platforms’, sabo ‘clogs’. If the listener has never encountered such designation before, then he most likely will not understand what is meant. In Danish, the very transparent structure of composite naming serves as an aid to the speaker in categorizing the general meaning. Cf. Rus. valenki and Dan. filtstøvler, Rus. krossovki and Dan. joggingsko (lit. ‘jogging shoes’), Russian tapki and Dan. hjemmesko, hyttesko (lit. ‘house-shoes’, ‘hut-shoes’), Rus. sandalii and Dan. (apart from sandaler) remsko (lit. ‘belt-shoe’), Rus. puanty and Dan. tåspidssko (lit. ‘big toe-tip-shoe’), Rus. platformy and Dan. plateausko (lit. platform-shoe), Rus. sabo and Dan. træsko (lit. ‘wood-shoe’).

An interesting remark was put forward by Erik Hansen, who in Sprogminuttet (radio broadcast devoted to the Danish language) discussed why in some rare cases in modern Danish the first component of a compound word seems redundant, cf. overfrakke = frakke ‘coat’, redekam = kam ‘comb’, borddug = dug ‘tablecloth’. Erik Hansen’s explanation was precisely to show that initially the first component of these compounds also had a specifying semantic difference, lost with the disappearance of the words associated with it. So, frakke meant ‘camisole’ and was opposed to overfrakke ‘streetwear’, borddug meant ‘fabric covering the table’ and was opposed to the words olmerdug ‘cloth for feather bed’ and sejldug ‘canvas for sails’, and the word redekam meant ‘comb for long hair’ and was opposed to nakkekam ‘hair comb’ (lit. ‘neck-comb’) – the name of now obsolete decoration (Hansen and Lund Citation1988).

Verification in the Backwards dictionary (Dansk Sprognævn Citation1988) shows that the more productive the compound key word, the stronger is its categorizing function. Thus, NDO registers 29 compounds with -sko as head noun; 10 with -tang as head: bidetang ‘nippers’, knibtang ‘pliers’, niptang ‘tweezers’, rørtang ‘adjustable wrench’, etc.; 28 with -bane: jernbane ‘railway’, kørebane ‘carriageway’, landingsbane ‘runway’, løbebane ‘running track’, fodboldbane ‘football field’, glidebane ‘skating rink.’, etc; 55 with -dyr as head noun: bløddyr ‘shellfish’ (lit. ‘soft animal’), dovendyr ‘sloth’ (lit. ‘lazy animal’), handyr ‘male’ (lit. ‘he-animal’), pattedyr ‘mammal’ (lit. ‘suck-animal’), rovdyr ‘predator’ (lit. ‘robbery-animal’), etc.

The compounds with -dyr demonstrate such high degree of generalization that it can be compared with the auxiliary function of the suffix component. Cf. the suffix denoting a baby animal -ling, usually considered productive, but registered in the NDO in this meaning in only four cases: kylling ‘chicken’, killing ‘kitten’, gæsling ‘gosling’, sperling ‘kind of small cod’ plus a dozen more words denoting, for instance, a child or a small, weak or beloved person, such as rolling ‘baby’, yndling ‘favorite pet or person’, svækling ‘weakling’, gamling ‘old man’. Compared with -ling, the 55 examples with -dyr can convincingly claim the status of a productive word-formation model.

3.4. Compound vs. suffixal categorization in Danish

Quite often, the lexico-grammatical categorization by means of the broad-meaning lexical component within the compound is based on the same principle as in the categorization by means of a suffix. Cf. for example, the ratio of broad-meaning verbal components -gøre ‘to do’ (40 examples in NDO) and -sætte “to set” (20 examples in NDO) and the suffix -ere, which have a similar lexical and grammatical meaning of causation, transfer into a certain state: Cf. blødgøre ‘soften’, færdiggøre ‘finish’, fastsætte ‘set’, igangsætte ‘set in motion’ = ‘activate’ and aktivere ‘activate’ = ‘set in motion’, etablere ‘establish’, pulverisere ‘pulverize’. The typical difference is that the borrowed suffix -ere is usually combined with borrowed stems, while the broad-meaning composite component is combined with native stems, which indicates that compounding as a way of generating lexico-grammatical meanings is the most adequate for the structure of the Danish language.

The ratio of suffixes for denoting abstract nouns is similar. Cf. the borrowed suffix -tion, which in addition to its main meaning, i.e., the naming of a process, is often used for categorization on the basis of “association of people engaged in a certain type of activity”: organisation, institution, inkvisition, inspektion (‘organization’, ‘institution’, ‘inquisition’, ‘inspection’). A similar meaning is formalized in Danish through a productive compound model with the second element -væsen ‘essence’/‘department’, used to name government organizations with different management functions: brandvæsen, postvæsen, retsvæsen, skattevæsen, skolevæsen, toldvæsen, vejvæsen, respectively, ‘fire, postal, legal, tax, school, customs, road department’.

Of particular note is the ratio of suffixes with the meaning of a person (male -er, -ar, -ør, -ent, etc. or female -erske, -inde) and words of the same semantics included in the compounds (-mand ‘man’ or -kone ‘wife’, ‘woman’, -pige ‘girl’, -dame ‘lady’, -kvinde ‘woman’).

There are 156 compounds registered in NDO with the word -mand. Compounds with the designation of females are less frequent, which, in fact, is also typical for the corresponding suffix designations: 13 compounds with dame were registered, 6 with kone, 6 with pige, and 4 with kvinde. The gender equality issue should also be mentioned here: many Danes in these cases nowadays insist on saying forperson (not formand) ‘chairman’, talsperson (not talsmand) ‘speaker’, and so on.

It is interesting to see how categorization develops using each component with broad semantics.

The stem -mand is used to refer to males. In this categorizing meaning, it is opposed to the stem -kvinde, denoting females: cf. tillidsmand and tillidskvinde ‘trade union representative – male or female’, formand and forkvinde ‘chairman’ (about a man and a woman), troldmand ‘sorcerer’ and troldkvinde ‘sorceress’. Such a correlation also repeats the not too numerous gender correlations of the suffixes -er and -erske: væver “weaver – male” and væverske ‘weaver – female’, arbejder ‘worker – male’ and arbejderske ‘worker – female’.

Other areas of categorization for the stem -mand are the designation of occupation (sømand ‘sailor’, brandmand ‘fireman’, forretningsmand ‘entrepreneur’), quality (levemand ‘bon vivant’, flovmand ‘mediocrity’, rigmand ‘rich man’), nationality (franskmand ‘Frenchman’, nordmand ‘Norwegian’). It is important to emphasize that exactly the same types of generalizing meanings can be expressed using the suffix -er: occupations as lærer ‘teacher’, dommer ‘judge’, quality as in drømmer ‘dreamer’, fanatiker ‘fanatic’, nationality as in tysker ‘German’, russer ‘Russian’.

Even paired compound and suffixal word-combinations with the same meaning are possible: gerningsmand and forbryder ‘criminal’, drabsmand and dræber ‘murderer’, tækkemand and tækker ‘roofer’. The main feature that distinguishes both models is that the -er suffix is combined primarily with verbal stems (although denominative formations are sometimes possible), and the -mand component is productive only in combination with nominal stems.

With all of the above, I want to show that composite stems with broad semantics are quite comparable to the most productive suffixes in the degree of generalization of lexico-grammatical categorization. At the same time, it is significant that Danish lexicologists consistently do not classify as suffixes those stems that can be used as separate words.

For instance, at the beginning of NDO, a complete list of Danish suffixes is given. An impressive number of lexically transparent components are listed there as suffixes, for example, such formants as:

-mager to designate craft professions (sadelmager ‘sadler’, skomager ‘shoemaker’), or with a derogatory connotation (projektmager ‘dreamer’, versemager ‘rhyme-spinner’); These words can also be used in more neutral or positive senses. For example, these days, projektmager can also be used as a (neutral/positive) job description, say, people who are good at raising funding for research projects.

-mageri to denote a trade or craft workshop (pottemageri ‘pottery/pottery workshop’, urmageri ‘watchmaking/watchmaking workshop’), or to denote a negative or tedious activity (fidusmageri ‘quackery’, kurmageri ‘courtship’);

-lænder from land ‘country’ to indicate nationality (englænder ‘English’, flamlænder ‘Flemish’);

-foldig from folde ‘to fold’; trefoldig ‘threefold’, syvfoldig ‘sevenfold’;

-holdig from holde ‘to contain’ - blyholdig ‘containing lead’, fedtholdig ‘containing fats’;

-cifret from cifre ‘digit’ sekscifret ‘six-digit’,·tocifret ‘two-digit’;

-etages from etage ‘floor’ femetages ‘five-storey’, tietages ‘ten-storey’;

-fodet from fod ‘foot’ barfodet ‘barefeet’, platfodet ’flat-footed’;

and many other bahuvrihi compounds – complex adjectives, derived from the names of body parts with the suffix -et: -hovedet mangehovedet ‘many-headed’; -hudet lyshudet ‘light-skinned’; -hændet hårdhændet ‘heavy-handed’; -mavet tykmavet ‘fat-bellied’, -skuldret bredskuldret ‘broad-shouldered’ and others.

Common to all of them is the impossibility to use the second component of the compound as an independent word. Apparently, this criterion was so significant for the NDO lexicographers that they decided to differentiate between the suffix and the stem not along the line “grammatical – lexical meaning”, but along the line “possibility – impossibility of separate functioning of an element”, which demonstrates the priority of analytical tendencies in the linguistic consciousness of Danes.

4. Conclusion

Summing up, we can conclude that in the field of Danish composite word formation, all the features of the analytical nomination identified by researchers for other Germanic languages and on other language levels are most consistently manifested. These are structural discreteness, potential distancing and modularity of components in the compound word, their semantic transparency and functional differentiation, as well as high productivity of the models they form. These are typical phenomena both in the sphere of nominal and verbal naming, which fully correspond to the communicative type of the Danish language as a hearer-oriented language.

Comparing Danish and English with Russian and French, we can see that exocentric languages (French and Russian) tend to be more reality-oriented, while endocentric languages (Danish and English) tend to be hearer-oriented. This highlights the obvious correlation between Michael Herslund’s concepts of “endocentric and exocentric languages” and Per Durst-Andersen’s concepts of “hearer-oriented and reality-oriented languages”.

Disclosure statement

No potential conflict of interest was reported by the author.


  • Baron, Irène. 2002. “Sprogtypologiske forskelle i oversættelsesperspektiv med særligt henblik på nominalenheder i dansk og fransk” [Language typological differences in translation perspective with particular regard to nominal units in Danish and French]. In Lingvistisk teori i oversættelsesstudier [Linguistic theory in translation studies], Copenhagen Working Papers in LSP (1), edited by Victor Smith, 3–23. København: Handelshøjskolen.
  • Baron, Irène, and Michael Herslund. 2005. “Mikrostruktur: ord og syntagmer. Leksikon” [Microstructure: Words and phrases. Lexicon]. In Sprogtypologi og oversættelse. Endocentriske og exocentriske sprog [Language typology and translation. Endocentric and exocentric languages], edited by Iørn Korzen and Lita Lundquist, Chap. 1.A, 18–39. København: Handelshøjskolen.
  • Diderichsen, Paul. 1976. Elementær dansk grammatik [Elementary Danish grammar]. 3rd ed. København: Gyldendal.
  • Durst-Andersen, Per. 1995. “Mental’naja grammatika i lingvističeskie suprtipy” [Mental grammar and linguistic supertypes]. Voprosy jazykoznanija 6: 30–42.
  • Durst-Andersen, Per. 1997. “Towards a Theory of Linguistic Supertypes: Speaker-Based, Hearer-Based and Reality-Based Languages.” In Proceedings of LP, edited by B. Palek, 125–152. Prague: Charles University Press.
  • Durst-Andersen, Per. 2011. Linguistic Supertypes. A Cognitive-Semiotic Theory of Human Communication. Berlin: Mouton de Gruyter. https://doi.org/10.1515/9783110253153.
  • Ginzburg, Rozalya, Sarra Khidekel, Galina Knyazeva, and Alexander Sankin. 1979. A Course in Modern English Lexicology. Moscow: Vysšaja Škola.
  • Guchman, Mirra. 1981. Istoricheskaya tipologiya i problema diahronicheskih konstant [Historical typology and the problem of diachronic constants]. Moscow: Nauka.
  • Hansen, Erik. 1996. Ping og pampersprog [The language of officials and bureaucrats]. København: Hans Reitzels Forlag.
  • Hansen, Erik, and Jørn Lund. 1988. 60 sprogminutter. Udvalg af artikler om dansk sprog og sprogbrug fra radioudsendelsen “Sprogminut” [60 language minutes. Selection of articles about Danish language and language usage from the radio broadcast “Sprogminut”]. København: Hans Reitzel.
  • Herslund, Michael. 1994. “Verbonominale konstruktioner og komposita” [Verbonominal constructions and compounds]. In NORDLEX-Projektet. Sammensatte substantiver i dansk [The NORDLEX project. Compound nouns in Danish]. LAMBDA 20, edited by Irène Baron, 83–88. København: Handelshøjskolen.
  • Herslund, Michael. 2000. “Tipologia grammaticale et tipologia lessicale” [Grammatical typology and lexical typology]. In Argomenti per una linguistica della traduzione [Topics for linguistics of translation], Gli Argomenti Umani 4, edited by Iørn Korzen and Carla Marello, 11–18. Alessandria: Edizioni dell’Orso.
  • Herslund, Michael. 2002. “Tre leksikaliseringsmodeller: de danske, franske og russiske bevægelsesverber” [Three lexicalization models: The Danish, French and Russian verbs of movement]. In Lingvistisk teori i oversættelsesstudier [Linguistic theory in translation studies]. Copenhagen Working Papers in LSP (1): 41–51.
  • Herslund, Michael, and Irène Baron. 2003. “Language as World View. Endocentric and Exocentric Representations of Reality.” In Language and Culture. Copenhagen Studies in Language 29, edited by Irène Baron, 29–42.
  • Ilyina, Tatiana. 2004. “Novye tendencii v affiksal’noj derivacii sovremennogo anglijskogo yazyka” [New trends in affixal derivation of modern English]. Bulletin of Novosibirsk State University 2: 54–57.
  • Jessen, Ida. 2005. ABC. København: Gyldendal.
  • Jessen, Ida. 2009. Azbuchnaya istoriya, Translated by N. Fedorova. Moscow: Tekst.
  • Korzen, Iørn, and Lita Lundquist, eds. 2005. Sprogtypologi og oversættelse. Endocentriske og exocentriske sprog [Language typology and translation. Endocentric and exocentric languages]. 3rd ed. København: Handelshøjskolen.
  • Krylova, Elvira. 2021. Kommunikativnye funktsii modalnyh chastits v datskom yazyke [Communicative functions of modal particles in Danish]. Moscow: MAKS Press.
  • Kuzmenko, Yuri. 1991. Fonologicheskaya evolyuciya germanskih yazykov [Phonological evolution of the Germanic languages]. Leningrad: Nauka.
  • Plotkin, Wulf. 1989. “Tipologicheskaya evolyuciya germanskih yazykov” [Typological evolution of the Germanic languages].” Lingvisticheskie edinicy, konstrukcii i tekst [Linguistic units, constructions and text], edited by Margarita Nikolaevna Vall, 3–13. Novosibirsk: Nauka.
  • Shaposhnikova, Irina. 1999. Sistemnye diahronicheskie izmeneniya leksiko-semanticheskogo koda anglijskogo yazyka v lingvo-etnicheskom aspekte. [Systemic diachronic changes in the lexico-semantic code of the English language in the linguo-ethnic aspect]. Doctoral diss., Irkutsk.
  • Smirnitsky, Alexander. 1956. Leksikologiya anglijskogo yazyka [Lexicology of the English language]. Moscow: Izdatel’stvo literatury na inostrannyh yazykah.
  • Smith, Viktor. 2005. “Andre indoeuropæiske sprog: tilfældet russisk” [Other Indo-European languages: The case of Russian]. In Sprogtypologi og oversættelse. Endocentriske og exocentriske sprog [Language typology and translation. Endocentric and exocentric languages], edited by Iørn Korzen and Lita Lundquist, 202–206. København: Handelshøjskolen.
  • BTSRYA –Kuznetsov, Sergey ed. 1998. Bol’shoj tolkovyj slovar’ russkogo yazyka [Big Explanatory dictionary of the Russian language]. Author’s edition 2014. Edited by St. Petersburg: Norprint. Accessed September 1, 2022. https://www.gramota.ru/slovari/dic/?word=%D1%80%D0%B0%D1%81%D0%BF%D0%B0%D1%85%D0BD%D1%83%D1%82%D1%8C&all=x.
  • Dansk Sprognævn [The Danish Language Council]. 1988. Baglænsordbog. Dansk sprognævns skrifter 15 [Papers of the Danish language council 15]. København: Gyldendal.
  • DDO – Asmussen, Jørg, Laurids Kristian Fahl, Jonas Jensen, Henrik Lorentzen, Sanni Nimb, Lars Trap-Jensen, and Thomas Troelsgård, eds. Den Danske Ordbog. Accessed September 1, 2022. https://ordnet.dk/ddo/ordbog.
  • COD – Thompson, Della, ed. 1995. The Concise Oxford Dictionary. 9th ed. Oxford: Oxford University Press.
  • NDO – Becker-Christensen, Christian, ed. 1996. Politikens Nudansk Ordbog. København: Politikens Forlag.
  • NSO – Lindgren, Birgitta, Helg Bruland, Alan Karker, and Ståle Løland, eds. 1994. Norsteds Skandinaviska Ordbok. Smedjebacken: Norsteds.

Reprints and Corporate Permissions

Please note: Selecting permissions does not provide access to the full text of the article, please see our help page How do I view content?

To request a reprint or corporate permissions for this article, please click on the relevant link below:

Academic Permissions

Please note: Selecting permissions does not provide access to the full text of the article, please see our help page How do I view content?

Obtain permissions instantly via Rightslink by clicking on the button below:

If you are unable to obtain permissions via Rightslink, please complete and submit this Permissions form. For more information, please visit our Permissions help page.