THE NORWEGIAN LEXICAL ITEM AKKURAT AND THE POLISH AKURAT : A COGNITIVE SEMANTIC ANALYSIS 1

The aim of the article is to demonstrate to what extent the Norwegian akkurat and the Polish akurat show similarities and diﬀerences in their conceptual content (meaning). Adopting the perspective of cognitive semantics (CS), as described in Langacker (1987) and Lakoﬀ (1987), I shall try to show that the meanings ascribed to these etymologically and formally related words constitute complex networks of senses, rooted in a prototypical centre in each of the languages under discussion. In addition to this, the ﬁndings will be interpreted with reference to the process of pragmaticalization (a language unit’s development of increasing pragmatic func-tions). Within this theoretical framework I shall demonstrate that subjectiﬁca-tion/intersubjectiﬁcation and pejoration/melioration motivate the main semantic diﬀerence between akkurat and akurat . The analysis is based on Norwegian and Polish monolingual corpus data.


Research context and aims
Akkurat and akurat share an etymological link as their roots are hypothesized to go back to the Latin verb accūr /o (-āre) 'apply care to' (Bańkowski, 2000, p. 7;Boye, 2005, p. 9;Steinnes and Vandvik, 1958, p. 10). What is more, they perform highly similar functions in Norwegian and Polish and seem to show clear similarities in meaning, with the result that several Norwegian conceptualizations can be directly mapped into Polish, and vice versa. Let us consider the following examples:  On the other hand, this kind of mapping is not always possible and it appears that akkurat and akurat are not infrequently to be seen as "false friends" in translation from Norwegian into Polish, or vice versa. So what is it that differentiates these lexical items in such cases? In this article I will take a stance on these issues. I shall try to show that the meanings ascribed to these etymologically and formally related words constitute complex networks of senses, rooted in a prototypical centre in each of the languages under discussion. Both categories will also be interpreted with reference to the process of pragmaticalization, understood to mean a tendency for a language unit to develop more abstract and pragmatic meanings (Defour et al., 2010, p. 168). Within this theoretical framework I shall demonstrate that subjectification/intersubjectification and pejoration/melioration motivate the main semantic difference between akkurat and akurat.
2 Semantic network and meaning construal in CS Cognitive semantics (CS) takes as its starting point the assumption that the structure of a category is based on the principle of family resemblance (Wittgenstein, 1953) and that it always encompasses central (prototypical) and peripheral members. Additionally, the boundaries of such a category are never clear-cut. They may shift depending on the available context and human judgment (Lakoff, 1987, p. 287). According to this view, a semantic category cannot be reduced to a list of features. It must be treated as a complex network of interrelated senses, which are contextually determined and created by language users (conceptualizers) for communicative purposes.
An important part of describing the meaning of a language item seems, then, to be distinguishing a prototypical centre. However, as Fillmore (1982) points out, such a reference point can be realized in many different ways. It can be identified, for instance, on the basis of "mutually compatible conditions", which constitute a conceptual content. The prototypical category members then represent all (or the majority) of such conditions that are excluded in less prototypical instances. On the other hand, one (or a few conditions) may possess a privileged status within the category. The members are then included into the network as derivative of this primary instance via the processes of metaphorical or metonymical extension. It should also be added that distinguishing a prototypical centre often proves to be impossible. In such cases the category membership is identified by approximating "an idealization of the category", which should be seen as a cooperation of conditions providing "cue validity" for it (Fillmore, 1982, p. 31-33). This kind of "idealization" is also important in Langacker's model of the category (Langacker, 1987, p. 16-17). He points out that the prototype-based category model is not sufficient for describing semantic structures. It should be complemented by using a schema, which is defined as "an abstract characterization that is fully compatible with all the members of the category it defines" (Langacker, 1987, p. 371). In such a model the category members are seen as nodes, which are linked to each other within the network by various categorizing relationships. The first type of these relationships is extension, fulfilled by the processes of metonymy and metaphor, e.g. from a global (or local) prototype. The second one is elaboration, which pertains to the relationship between a schema and its instantiations and which is connected with the processes of generalization or specification (conceiving an entity with finer detail). The third type of the relationships is mutual similarity (Langacker, 1987, p. 378f).
However, the cognitive model presented above for describing a semantic structure should not be taken too literally. It is rather to be seen as an idealized module and some kind of icon. According to Dirven and Ruiz de Mendoza Ibáñez (2010, p. 47), we should not think that language items possess "clearly defined different meanings". As these researchers put it, " [w]hat is at stake is rather that a given word form is applied in a number of different contexts of use and thereby exploits one or several of the [. . . ] processes of semantic extension". The same pertains to semantic networks, which are distinguished and described by linguists. They are not to be treated as mental representations of polysemy in our minds either. They are purely "linguists' representations of motivated extensions of form-meaning parings" (Dirven and Ruiz de Mendoza Ibáñez, 2010, p. 49). As Langacker points out: "a schematic network is a set of cognitive routines, entrenched to varying degrees: despite our inevitable reifications, it is not something a speaker has, but rather what he does" (Langacker, 1987, p. 382;italics original).
Within the framework of CS, meaning is also equated with conceptualization rooted in cognitive processing (Langacker, 1987, p. 97). According to this view the semantic structure (predication) does not exist objectively but is based on imagery, which is defined as humans' ability to construe a situation in different ways by selecting various features of it for explicit attention (Langacker, 1987, p. 110). The linguistic way of construing content is then connected with profiling, i.e. highlighting some structures within a conceptual base that is necessary for these structures' characterization (Langacker, 1987, p. 118). Taking this into consideration it can be concluded that the meaning of a language item always emerges from an interaction between the profile and the base. In addition to this, the encyclopaedic and context-dependent character of such interaction should be stressed, which implies that a semantic analysis should take into account not only the lexicogrammatical content but also pragmatic aspects of communication and various knowledge systems in discourse. The semantic content of an expression is simply negotiated by language users in each communication act (Langacker, 1987, p. 157).

Data and methodology
The data for this study was drawn from synchronic corpora -for Norwegian this was the Corpus for Bokmål Lexicography (LBK) with a total of 50,000,000 words, containing texts from 1985 until now, and for Polish the Polish National Corpus (NKJP) with a total of 1,500,000,000 words, containing selected texts from 1895-2010 and a demonstration version of the PWN Polish Corpus (PWN) with 7,500,000 words, containing selected texts from the period 1920-2000. Norwegian and Polish monolingual dictionaries were also occasionally consulted. As I did not have corpus collections containing naturally occurring Polish spoken data at my disposal, only written language corpora were used for data analysis in both languages. However, this drawback was minimized by the fact that the materials gathered in the mentioned corpora can be characterized as varied, encompassing texts belonging to a range of genres (e.g. classic literature, daily newspapers and specialist journals publications, advertising leaflets and also speech-related texts).
From these corpora, examples of the lexemes akkurat and akurat were extracted, each with considerable context -approximately 5000 examples for the Norwegian item and 2533 observations for the Polish word. The difference in the quantity of the analyzed conceptualizations between Norwegian and Polish was due to the Polish word's lower frequency in the corpus data. In the analysis, however, all available Polish examples from the full NKJP corpus (1800M segments), the balanced NKJP subcorpus (300M segments) and the demonstration version of the PWN Polish Corpus were taken into consideration. It should be added that the lower frequency of akurat in the Polish data may be explained by the fact that this item is clearly preferred in spoken Polish discourse.
In a first stage the occurrences of akkurat and akurat were subdivided into major types distinguished by observing distributional facets of the two items' usage. This analysis took the following factors into consideration: -The item's relation to a chosen unit in the sentence (related/not related) Akkurat/akurat is related to a chosen unit: e.g. Hvorfor akkurat nå 'Why right now', the chosen unit is the adverb nå 'now', which is modified; Akkurat/akurat is not related to a chosen unit: e.g. Vi er akkurat kommet fram 'We have just arrived', the meaning of the whole sentence is modified.
-The item's position into the chosen unit (pre-/post position) Akkurat/akurat is pre-positioned into the chosen unit: e.g. Akurat to nie powinno nikogo dziwić 'Particularly this should not surprise anyone', the chosen unit is the demonstrative pronaun to 'this'; Akkurat/akurat is post-positioned into the chosen unit: e.g. To akurat nie powinno nikogo dziwić 'This in particular should not surprise anyone'.
In some examples the more unrestrained word order in Polish caused problems with establishing the relationship between sentence components. In verifying what unit akurat relates to (e.g. if it is pre-or post-positioned), the most natural reading was then chosen. Table 1 summarizes the relative number of occurrences of the two items' major types of use. Within such major types of uses, semantic analysis was conducted to distinguish the more detailed content of akkurat and akurat. Not only semantic-pragmatic variables but also more formal features were analyzed. This concentrated mainly on three factors: -The item's obligatoriness in the sentence. Obligatory uses were considered to be those whose avoidance would make the sentence incomplete (grammatically or semantically) and thus unable to function in discourse, whereas a non-obligatory item only inputs additional information into discourse and can be omitted; -The item's occurrence in a particular kind of construction. Major attention was paid to structures, which were prominent in the data in each language. However, systematic analysis to describe the items' occurrence in different types of constructions was not undertaken and this topic needs further research; -The source of the discourse content to which the item refers (the conceptualizer's own words / somebody else's words); -The item's position in the sentence (initial/final). Table 2 gives an overview of the classification of akkurat and akurat into more detailed semantic types, which will be discussed in detail in the remainder of the article. After these introductory explanations, let us now pass on to the analysis of the linguistic material and look in more detail at the similarities and differences between the Norwegian akkurat and the Polish akurat at the level of organizing conceptual content. According to Langacker (1987, p. 370), semantic investigation not only "requires the listing of all conventionally established values of a lexical item [. . . ]. It further demands an analysis of how the category is structured, i.e. how the different senses are related to each other".
4 Akkurat/akurat related to a chosen unit in the sentence

Pre-positioned intensifying-contrasting modifier
The main function of the Norwegian akkurat is to focus the addressee's attention on an object, which is profiled by a chosen language unit (substantive, adverb, noun phrase, clause, etc.) in the sentence. It is fulfilled by highlighting the information that this object is exactly/just what the unit communicates. In this way the object becomes a figure and is more clearly distinguished from the ground. On the other hand, akkurat serves to contrast the chosen object with an another potential object, evoked in the conceptualization's base. This meaning of akkurat can be conceived as 'exactly/just X (and not Y)'. The item is pre-positioned onto the chosen unit in such cases. Examples 4 and 5 are a clear illustration of this content.
The same characteristics can be applied to the Polish akurat in examples 6 and 7. The item is here used as a pre-positioned intensifying-contrasting modifier, which highlights the content of the chosen unit in the sentence.
The intensifying-contrasting function of akkurat or akurat becomes clearly visible when the items are removed from the conceptualizations, which is potentially possible, e.g.: (8) Hvorfor nå? Why now? (9) Dlaczego ja?
Why me?

Pre-positioned intensifying modifier
Akkurat in conceptualizations 4, 5 above may be seen as close in value to those in 10, 11 below. In both cases the Norwegian item is used as a pre-positioned modifier, which intensifies the meaning of the chosen (described) unit in the sentence. Yet, while 4 and 5 give the impression that contrast is an important part of the linguistic image, in 10 and 11 this information is not even taken into consideration.
(10) Du har akkurat samme dialekt som moren min. (Rysst and Daren, 2007, p. 9) You have exactly the same dialect as my mum. (13) Cha! Cha! Cha! dbam ja o to wszystko akurat jak o śnieg zeszłoroczny. (Doroszewski, 1958(Doroszewski, -1962 Ha ha ha! I care about all this just as about the last year's snow. When expressing such content, the Norwegian akkurat frequently precedes a unit elaborated by a comparative clause (or phrase) introduced by the subordinate conjunction som 'as/like', as in example 11. In the Norwegian data, 1065 examples of this comparison construction were found. This construction does have its equivalent in Polish -example 13 -but it is used only marginally in this language. As the overview in Table 2 shows, such a use of akurat was found only twice in the whole Polish data.
As the function of contrasting is reduced in conceptualizations 10-13, the content of akkurat/akurat described in this Section can be conceived as 'exactly/just X'.
However, it should be pointed out that the strength of the contrast that is profiled by these items in conceptualizations 4-12 is clearly context-dependent. It depends also on the character of the object itself. For that reason the meanings characterised in Sections 4.1 and 4.2 are not considered separately in the overview in Table 2.

Post-positioned modifier
A characteristic feature of the Polish language is that it enables a more unrestrained use of akurat in its pre-and post-positional function, which, as we can expect, influences the conceptualization content (cf. Kurkowska, 1974, p. 212f). The intensifying-contrasting function becomes weakened when the Polish item is placed after the chosen unit, as illus-

Contrast within a construction
According to CS the meaning associated with a grammatical construction is an integral part of the described word's semantics. This kind of grammatical semantics can be demonstrated by our next extension of the category. In Norwegian, contrast can be highlighted by akkurat post-positional to the chosen unit. However, this seems to be possible nearly exclusively in conceptualizations including negation and implying contrast within its scope. This kind of construction-dependant content of akkurat is illustrated in example 18.
(18) Filmen ble ingen verdensslager akkurat, men fikk gode kritikker og ble en pen sucess i USA. (LBK) The film was no world hit exactly, but received acclaim and was a pretty success in the United States.
A similar construction can be also found in Polish, as example 19 demonstrates.
In both languages, however, this construction is very rare -only 3 examples of it were found in the Polish corpus data and 9 in Norwegian.
5 Akkurat/akurat not related to a chosen unit in the sentence An important condition that facilitates the widening of the categories is the possibility to use akkurat and akurat as items that are not related to a particular, chosen unit in the sentence. In such cases their main conceptual content, paraphrased above as 'exactly/just' (Section 4.1 and 4.2), becomes more independent and can potentially be mapped into varied abstract domains enabling the items to express new meanings. Additionally, the items' discursive potential usually increases in these kinds of conceptualizations. In such cases akkurat/akurat become intensifying/discursive operators that power the integration of the sentence into information flow in discourse by focusing on different kinds of information (Sections 5.1-5.6).

Temporal operator
In its temporal function the Norwegian akkurat can express a current point in time.
Example 20 is a clear illustration of this temporal meaning. However, the item's function here is to intensify not only the moment in time but also the result of the profiled verb action (er kommet).
The overview in Table 2, presented in the initial part of this article, shows that this content is also frequently expressed in Polish -as in example 21.

Contrasting operator
In Polish akurat can profile contrast not only by distinguishing a particular object in discourse (Section 4.1) but also by modifying the content of the whole sentence, which is shown in example 22. In such cases the item may be classified as a contrasting operator whose use in the sentence is optional. However, removing akurat from 22 would make the contrast disappear from the conceptual picture.
(22) Przebudowaliśmy nasze mieszkanie, bo akurat mamy skończone studia architektoniczne, ale nie chcemy zachęcać do tego innych. (NKJP) We have rebuilt our apartment because we [differently from other people] have completed architectural studies, but we do not want to encourage others to do so.
As we can see from the overview in Table 2, this content was frequent in the Polish data. However, it should be stressed that it was not always possible to classify such realizations of akurat in a relatively unequivocal way. The border between the item's contrasting and temporal uses -especially when a less precise time range was profiled -seemed blurred in several cases. 52 examples of such meanings were identified in the data and they were not counted in the overview in Table 2. Their conceptual content needs to be studied in more detail.
The content construal presented in example 22 is not conventionally allowed in Norwegian, where different ways of expressing contrast by akkurat are preferred (Sections 4.3.1 and 5.4). Rysst and Daren (2007, p. 9) point out that the Norwegian akkurat is frequently used to express the speaker's support or acceptance towards a content that has been previously articulated in discourse. In such situations the speaker's positive attitude to what has been said is also marked. This extension of the category can then be characterized as the realization of the item's intensifying function on a higher (more abstract) level in discourse. However, our analysis of the data showed that the affirmative akkurat was used in two types of situations: when the speaker related to his/her own words (cf. 5.3.1) or when s/he expressed support for what the somebody else is saying (cf. 5.3.2). Additionally, these functions of the item seemed to correlate with its syntactic position.

Emphasis operator
In its use as modifier, akkurat highlights the chosen unit's content in a sentence (cf. 4.1, 4.2, 4.3). What is highlighted in 23 is the content expressed by the speaker him-/herself in the preceding sentence or sentences. By confirming this content s/he focuses the addressee's attention on it. In such cases the Norwegian item is not obligatory and it mainly takes final position in the sentence (the item was placed in sentence final position in 23 of 27 identified examples of this use of akkurat). Its affirmative content is clearly emphatic.
(23) Det er et godt graffitiminne. Men en spennende historie å fortelle videre, er det jo ikke, akkurat. (LBK) It's a good graffiti memory. But an exciting story to tell to others, it is not, indeed.

Agreement operator
In the examples below akkurat occurs in utterance-initial position. Additionally, it is always obligatory, i.e. its omission would make the discourse structure incomplete. The profiled content varies here from the speaker's agreement towards the previously expressed proposition, as in example 24, to a yes-answer to the interlocutor's question, as in 25. None of these affirmative meanings are expressed by the Polish akurat nowadays. However, a kind of affirmative content (paraphrased as A jakże! 'Yes of course' Zapewne! 'Certainly') is present in an archaic use of this item, as in example 26 (Karłowicz et al., 1900, p. 20).

Contrariety fostering operator in the context of negation
As a means of expressing the speaker's reaction towards a content previously articulated in discourse the Norwegian akkurat occurs also in the context of negation, as in example 27. In such cases akkurat opens a space for a contrary reading of the negated anticipatory content. This reading can potentially be expressed in the following discourse. Thus, the sentence's integration into the information flow in discourse is here also driven forward.
(27) Slaraffenliv? Ikke akkurat! 07:00 : STÅ OPP! (LBK) A wonderful life? Not exactly! 07:00 : GET UP! Omitting akkurat from the conceptualization above would make the addressee not expect a continuation of the utterance including a content that may seem contradictory to that profiled in the preceding discourse. This use of the Norwegian akkurat, even if not frequent in the data (see overview in Table 2), shows clear similarity to the item's function as post-positioned modifier profiling contrast within the construction discussed in Section 4.3.1. However, the contrariety that is profiled by the construction in example 18 (ikke X akkurat, but . . . 'not X exactly, but. . . ') is in example 27 placed into the conceptualization's base.
The same characteristics can be applied to the Polish akurat. In example 28 this item co-occurs with negation, giving the addressee a hint that in the following discourse a content that is contradictory to what has been articulated previously may be expressed. This use of the Polish akurat may be seen as a basis for developing the item's function as contradiction operator (cf. Section 5.5).

Contradiction operator
Contrary to the Norwegian item the Polish akurat has acquired a function of contradicting the information provided in the former discourse. In such cases the item is characterized as an expressive adverb ("przysłówek o charakterze ekspresywnym"), by which the speaker can articulate his/her subjective attitude to what is being said (Szymczak, 1992, p. 28).
In example 29 the previously expressed content is denied.
In example 30 the item indicates the speaker's objection to the communicated idea.
By using akurat the speaker can also expresses his/her doubt and mistrust toward the articulated content, as in example 31.
Additionally, the item can profile disapproval of previously expressed expectations, as in example 32. The common denominator of these conceptualizations is that akurat is here used as an obligatory interaction indicator pronounced with the distinctive typically falling intonation, which signals the speaker's negative attitude to the previously expressed content. By using this item the speaker reacts to what the addressee said. These kinds of contents have not been identified in the Norwegian data.

Presentation operator
According to Rysst and Daren (2007, p. 9) the Norwegian akkurat "is often used in conversation when one party is [. . . ] bringing new information" into the discourse space. This clearly discoursal function of the item is demonstrated in example 33. Only two occurrences of this use is mentioned in the data (cf. This semantic-pragmatic function has not been developed by the Polish item.

Content 'as much as needed'
The next extension within the category expressed by the Norwegian akkurat can be schematically conceived as 'as much as needed ' -example 34. In such cases the item is an integral part of the sentence and cannot be omitted.
Example 35 demonstrates that a similar fully lexicalized node can be distinguished within the Polish category: (35) -O dwieście złotych. Będzie akurat. (NKJP) -About two hundred zlotys more. It will be as much as is needed.

Conclusion
In this study the semantic structures of two words -the Norwegian akkurat and the Polish akurat -were examined. We will now reconstruct the structures of the semantic networks developed by these items and discuss the similarities and differences between them. Figure 1 below shows the semantic overlap of radial categories created for akkurat and akurat. The extension relationship within the category is symbolized by a solid arrow, elaboration by a dashed arrow and the mutual similarity relationship by two dashed arrows (← − →). Conceptual distance between senses is also graphically represented. Additionally, a different colour has been chosen for each language. The more present a meaning is in a language, the more dominant the colour symbolizing this language (see also overview in Table 2   Taking this into consideration, akkurat and akurat in their use as an intensifying-contrasting modifier can be treated as the main prototypical reference point within the category. This node represents the majority of attributes (conditions) that enable the natural widening of the category, and as such it is the most logical centre of such conceptual operations. In this context the node [4.2], even though it has the widest applicability as a reference point in both languages, is to be seen as a more local prototype derived from [4.1]. It should be mentioned, however, that neither of the criteria used above can be considered as sufficient proof of prototypicality. It is obvious that additional research could be undertaken to complement the characteristics given here.
Other nodes within the category structured for akkurat and akurat can be seen as metonymical extensions from this complex prototypical reference point. However, one point should be made here. The central content of the item(s), which was articulated in [4.1] and [4.2] and which pertains to its adnominal function, can become independent and also be expressed by akkurat/akurat when unrelated to a chosen unit in discourse. This warrants the development of the nodes in which the items are used as operators. From this matrix content, paraphrased as 'exactly/just', several extensions of the category have developed via mapping into more abstract domains -[5.1], [5.2], [5.3], [5.7]. A clear example of that is the temporal content of the items -[5.1]. Analyzing the meaning of akkurat/akurat within the network category, we may explain the existence of this content as derived from the items' use as intensifying modifiers within the phrase '(in) exactly/just this moment in time'. While focusing the addressee's attention on a particular moment in time akkurat/akurat can profile this content autonomously by putting the modified head noun's content into the conceptualization base (i.e. by implying this content). This moment in time can be then broadened into a less precise (delimited) time range, which is highlighted in some contexts. In such cases conceptual similarity between the Polish item's temporal [5.1] and contrasting meaning [5.2] seems to be a fact, as distinguishing these contents was difficult in several conceptualizations. However, this issue must be studied more thoroughly.
The semantic functions of the described items also correlate with their syntactic environment, which should be seen as the next important mechanism underlying the categorybuilding. The word's post-position in relation to the chosen unit in the sentence causes weakening of its contrasting and intensifying function, but on the other hand, it may also amplify contrast within a particular construction [4.3.1] -something which demonstrates itself particularly clearly in Norwegian. Deriving [4.3.1] from [4.1] is then more plausible than the alternative. Another syntactic mechanism that clearly motivates the derivation of [5.3], for instance, relates to the items' use as pre-positioned modifiers relating to a comparative unit (a phrase or a clause). This kind of comparison construction, illustrated in examples 11 and 13, is more frequent in Norwegian (see the overview in Table 2) but does, exist also in Polish. It can be seen as the structure that enables further growth of the category into affirmative functions. The relation of comparison can be applied to new contexts, e.g. akkurat nå 'just now' → akurat som i 1914 'just as in 1914' → akkurat som i alle andre siviliserte land 'just as in all other civilized countries', etc. In example 36 below this relation is conceived as 'akkurat som Peter ville gjort' 'exactly as Peter wanted to have it done'. My cousin went there ten years ago or more.
In this context the widening [5.7] is also motivated -'exactly/just [as is needed]'. A further syntactic condition that enables category growth involves position in the sentence. The opposition between the item's initial and final position is relevant for distinguishing different meanings within the node of affirmative contents in Norwegian -[5.3]. The initial (thematic) position in the sentence is also crucial for establishing the Norwegian item's function as a presentation operator [5.6]. In conceptualizations representing this node the scene construal is cognitively motivated. In order to introduce a new element into the discourse space, it is good to prepare the ground for the addressee. Akkurat in the initial position in the sentence is a useful tool to do so, as the item resembles akkurat in its affirmative function as agreement operator [5.3.2]. The content that is affirmed, however, is here only implied vaguely and the speaker is pretending that it is accessible to the addressee (due to sentence initial position). In this way a new node can develop. The diagrams of the Norwegian and Polish categories, which are conflated in Figure 1 (p. 232) can also shed some light on the question of other, more general, mechanisms underlying motivation for the particular instances of akkurat or akurat. For instance, the comparison construction including akkurat may be seen as very productive within the Norwegian category.  Figure 1 Semantic networks of lexical items akkurat and akurat ral indications for a prototypical structure of the categories expressed by akkurat and akurat. Figure 1 above also demonstrates when the potential compatibility of nodes in Norwegian and Polish is highest and when it is lowest. However, such interchangeability should be proved in a bilingual corpus, which is a subject for further research. Generally speaking the similarity between the items is most noticeable within the nodes conceived as The Norwegian and Polish categories can be also interpreted in more general terms within the framework of traditional typology of semantic changes (cf. Hollmann, 2009, p. 527). The first traditional classification in meaning concerns whether the semantic structure becomes broader or narrower. The tendency which can be observed here definitely is to be characterized as a widening of the scope of the category (the item's content) in both cases (Hollmann, 2009, p. 528). The broadening of the word's scope of meaning can be treated as an index of the underlying process of pragmaticalization (Defour et al., 2010, p. 168), which means a tendency whereby a language unit develops more abstract and pragmatic meanings. Such a unit is gradually becoming a so-called pragmatic marker, which, as Aijmer et al. (2004Aijmer et al. ( , p. 1783) point out, "explicitly indicate[s] the speakers' awareness of the communication process as taking place in a context and thereby help to shape that process in a particular way".
Pragmatic markers are used to negotiate meaning between the speakers in the process of communication and such negotiation involves the continual updating of several (explicit or implicit) assumptions, e.g. by challenging them or denying them. So, markers of this kind are necessary in order to constrain or guide the interpretation process in discourse (Aijmer et al., 2004(Aijmer et al., , p. 1784).
As we have seen, both the Norwegian akkurat and the Polish akurat fit this description. They can, for instance, strengthen a contrast implied within the sentence. Additionally, both words show category growth from "more concrete" meanings -e.g. [4.2], [5.1] -to clearly abstract encodings of the speaker's judgment respecting conditions of evaluation -e.g. [5.4], [5.5], [5.6]. They can signal disagreement with a previous claim or confirm what has been said. Additionally, the Norwegian item serves to express emphasis and it seems to be developing the function of a presentation operator, which is used to introduce new, but contextually grounded information into discourse.
In addition to this, the pragmaticalization of the category can be characterized in finer detail. It can be described as a shift towards meanings that express the perspective of the speaker, which in the literature is called subjectification (cf. e.g. Traugott, 1995, p. 31-32 for a thorough description of the concept). Hollmann (2009, p. 536) suggests that "this regularity may be related to what we may call certain egocentric tendencies of human cognition". As we have demonstrated, both items' meaning is gradually becoming more subjective.
The conducted analysis showed, however, that this tendency seems to be predominant in Polish, where the item akurat has developed several interrelated contents as a contradiction operator. We find increased subjectification in conceptualizations focused on negation of, disbelief towards, and objection to a previously expressed content, which are typical for this item. In such cases the speaker primarily expresses his/her attitude to what has been said in discourse.
On the other hand pragmaticalization can be connected with a shift towards meanings that in discourse involve increased attention to the addressee, which in cognitively influenced semantics goes under the name intersubjectification (Defour et al., 2010, p. 168). In this respect the Norwegian akkurat seems to be more representative.
As a presentation operator [5.6] this item expresses mainly addressee-oriented content. Such uses of akkurat can be characterized as interactive to the extent that they serve the speaker to guide the hearer in the interpretation of the content, which is created in discourse. In this way the speaker seems to take responsibility for the success of the communication, as s/he is taking into consideration the addressee's attempts to integrate the new information. S/he is trying to be more cooperative in language communication.
The second traditional classification in meaning concerns the preferred direction in the development of the item's conceptual content. In Norwegian we can observe a clear tendency towards what is called melioration, i.e. the developing a more positive meaning of akkurat. This mainly pertains to the position that the affirmative contents [5.3.1] and [5.3.2] possess within the category built by the Norwegian item, which the Polish items shows only marginally.
In Polish, on the other hand, a path of a more negative meaning is developing, which is known as pejoration. It should be clear from the preceding discussion that in this language the item's contradiction function (implying discredit and disapproval highlighted by the falling intonation) is becoming dominant. The related contents created on this basis can be seen as indicators of the speaker's more negative attitude to an object, which s/he expresses in discourse.
The tendency to develop meanings, which are based on subjectification or intersubjectification, and pejoration or melioration, can be seen as the main semantic difference between the Norwegian and Polish item. However, what deserves further study is whether this tendency can be demonstrated in spontaneous conversation in both languages, i.e. when the possibility of creating pragmatically motivated contents increases. This, at any rate, can be examined in future research.