The paper addresses adverb just in Middle and Early Modern Corpora, which is classified as a focusing restrictive employed to delineate the focus value more emphatically without explicit highlight of alternative values [1, p. 35]. As a focus marker in Modern English just can be used as an exclusive meaning "only" (1) or as a particularizer in the sense of "exactly" (2) appearing in front of the syntactic constituent it modifies. E.g. COCA search data [2] reveal three meanings typical of just in Present-Day English (PDE): 1) focusing exclusiveand nothing more; focusing 2) particularizerexactness or preciseness; 3) temporal adverbial with the meaning a moment ago, highlighting that the adverb is less likely to occur in academic style, while it's more applicable in TV discourse (Fig. 1). The current research is aimed at identifying grammaticalization cline for Middle and Early Modern English focusing adverb just as an information-structural marker in Present-Day English. The scientific novelty of the study is connected with applying information-structural analysis to Corpus data retried based on Discourse Representation Theory [3; 4] and Cesax software model [5; 6], which allows automated and semi-automated marking of information-structural components on the basis of automated and semiautomated algorithms. This will allow providing a better insight into quantitative and qualitative data interpretation in terms of grammaticalization processes, as well as exemplifying correlation of focusing adverbs standing and development and their role in the information structure of the sentence from the diachronic perspective.
The category of adverbs represents a wide range of meanings starting with identifying time to location or manner. In the process of language development adverbs can be differentiated by their degree of grammaticalization (special suffixes that emerged as a result of lexicalization or desemantization) [7, p. 508]. Adverbs generally represent the most heterogenous class of lexicon, which served a "wastepaper basket" for all the words that did not fit the main categories, viz. nouns, verbs, adjectives [8; 9]. Focusing adverbs as a class was singled out at the end of the XX-century-grammars. In modern grammars the other terms are applicable when referring to this class: "focusing modifiers" [10; 11; 12], "focusing particles" [13; 14; 15], "scalar operators" [16]. The studies of focusing adverbs in modern linguistics [14; 16; 17; 18; 19; 20; 21] allow singling out restrictive and additive adverbs with their further division into exclusives, particularizers, scalar and non-scalar additives.
Historically the mentioned above adverb goes back to Latin adjective iustus and adverb iust that had an equivalent in Middle French in the form of the adverb justement introduced into English firstly as an adjective by the end of the XIV cen. [22]. Corpus studies of Middle and Early Modern English adverb just can shed light on the questions of its grammaticalization, as well as, its transformation into a Focus marker throughout XIV-XVII cen. English.
In the light of the above, the current research explores adverb just in Corpus of Middle English Prose and Verse (ME Corpus) and Early English Books Online Corpus (EEBO Corpus), where presumably its major adverbial function is focusing particularizer meaning 'exactly, precisely' [23]. Additionally, according to E. Traugott [24] just in Middle English (ME) was employed as a manner adverb rendering the meaning of 'fittingly', however, this function did not survive in PDE. E.g. Corpus data in the current studies are aimed at checking the following hypothesis: the cline for adverb just grammaticalization in English presupposes its initial functioning as an ADJECTIVE → POLYSEMOUS ADVERB → PARTICULARIZER →? EXCLUSIVE ADVERB. As K. Aijmer [25] speculates polysemous meaning of just arises as a result of implicature or inference licensed by conversational principles, albeit the researcher agrees with T. Nevalainen [23, p. 16] that during earlier stages of language change such transformations are possible due to metaphorical abstractions. From the perspective of the current study the meaning expansion and further specialization correlates with syntactic changes within the period under investigation with adverbial function arising gradually within the period of general word order (WO) normalization that led to finding new means in the language to represent information-structural relations [26]. Considering this fact, another aspect to be taken into account is connected with variations in just positioning in the sentence due to its general dependency on Focus as a defining characteristic of focusing adverbs [14; 27; 28]. Such complex approach enables to define the correlation of the suggested grammaticalization cline, syntactic peculiarities and information structure marking (IS) in the course of XIV-XVII-century English development.
To observe the regularities of WO fluctuations in tokens with just in two Corpora under analysis a special attention is given to different actualization statuses of information (given-new), as well as, Foci types marked by the adverb. The theoretical framework for given-new annotation is based on Discourse Representation Theory [3; 4; 29; 30; 31; 32] eliciting several types of contexts in the text fragments: discourse, encyclopedic, situational and scenario-context. To introduce the theoretical framework for the investigation, I'll recap some peculiar feature of discourse analysis, as well as, present the coreference editor that partially allows automated and semi-automated coding of the givenness status among discourse referents.
Discourse Context. The study of givenness suggests building a file that retains all the DPs previously mentioned in the text, which are known as discourse referents (DRs). The core idea of the framework is to track whether the DP is a new DR or not. To illustrate the principles of givenness annotation Haugh, Eckhoff & Welo [29] demonstrate how the hearer or reader frames up a mental model of the current discourse, identified as a discourse representation structure (DRS). It consists of two parts: a universe of DRs and a range of DRS criteria that help encode the information. In DRT the box visualization of the sentence in ( 4) is assigned the manifestation in Fig 2 . 4. A wolf howled. Suppose that sentence (4) is followed by sentence (5).
The discourse context representation of sentence ( 5) is illustrated in Figure 3. Pronominal it shown as a DR y corresponds to a discourse referent highlighted in the previous context. Hence, the next operation for sentences (4) and (5) termed merging of DRS is given in Fig. 4.
Another case is Encyclopedic context, which allows to level those discourse referents that have not been highlighted in the previous discourse though they may be perceived as given for the reader (king of jews instead of Jesu). E.g.
Situational context allows to level the discourse referents based on the general information. E.g. 9. This man is a robber. Based on this, each discourse referent is tagged on the given-new plane taking into account the extended annotation scheme [30; 33] (Table 1).
Assignment of tags in the extended annotation scheme for IS
Labelling all the relevant constituents for their information status presupposes employing "coreference resolution" [34], which shows whether DP refers back to another element and if so, with what type of link (active, accessible, inferable, non-specific). The computer software that allows DP annotation of the data is called "Cesax". It enables to resolve coreference semi-automatically presupposing that the automatic search is enriched with the possibility to ask the user input in ambiguous cases. Currently the list of DP features is represented by the three members [6], e.g.
The search algorithm for new-given information detection is described in Komen [5; 34] and is summarized in 11.
With the commencement of coreference resolution process, the software automatically determines as many anaphoric links as possible, as well as, referentially new DPs (lower-right yellow window in Fig. 9) [6]. However, once a solution corresponds to one of the built-in suspicious situations, Cesax stops to ask for the user's decision. After labelling given-new information in the discourse, the informational status of individual linguistic expressions is further identified in terms of Topic and Focus marking. For the current study Topic and Focus are identified in line with Krifka's [35] and Reinhart's [36] definitions. Thus, the former highlights the information the sentence is about (represented by aboutness, given, familiar and contrastive subtypes), whereas focus is associated with salient or the most important information in the sentence [37, p. 143]. It is further subdivided into informational (a sentence element that stands for a great level of novelty) [38], identificational (refers to the presence of alternatives prior available in the discourse) [39], emphatic (represents the elements that demonstrate the extreme value on the scale of values) [40], exhaustive (renders the exclusion by identification with respect to the alternative propositions), contrastive (the components of the common ground that contain a proposition the sentence can be contrasted against) [41], verum (the truth value of the sentence) [42] and mirative (surprising or unexpected information) [43] Foci. Table 2 summarizes the tags applied to the analysis the second type of dichotomy, viz. Topic/Focus.
The results of tagging the components are given in Figure 10. The abovementioned methodology allows complex tagging of clause components in parsed corpora in regards to information givenness, as well as, identify sentence Topic and Focus and their variations. Such analysis enables to identify the role of a particular focusing adverb when marking sentence Focus or Topic, define its role in correlation of IS and sentence word-order, as well as, explain the grammaticalization mechanism on a specific historical language layer, which provides new quantitative and qualitative insights into the English language development.
As it has been previously highlighted, the patterns with adverb just and its spelling variants are retrieved from two diachronic Corpora, which differ in terms of their size, lemmatization and tagging. Corpus of Middle English Prose and Verse is represented by 300 Middle English primary texts collection marked using basic TEL semantics and available for bulk download in XML form [44]. It allows conducting a basic search, proximity search, citation search, etc. (See Fig. 11). This type of search pinpoints the emergence of a specific linguistic phenomenon in the language, as well as, calculates its frequency in a certain period of time. However, further semantic and syntactic analysis of the linguistic form requires manual processing based on peculiar features of a clause and XP structure in order to separate, in our case, adverbial just (ME forms iust(e), just(e)) from its nominal, adjectival and verbal counterparts.
Early Modern English Corpus EEBO developed by the Text Creation Partnership contains ca. 755 million words in 25,368 texts within the time frame of 1470s-1690s. It has a lemmatization and part of speech tagging, thus, simplifying the query process [45]. The search intends to provide the raw data for two Early Modern English spelling variants just and iust.
After automated search all the examples with NPs and APs containing just are tagged manually (since automated search sporadically fails to differentiate just as an adverb or an adjective), which can be observed after conducting the automatic comparative analysis of the collocates (Figure 13). It should be noted that such collocates as just judgement, just fears, iust law demonstrate that the word just functions as an adjective rather than an adverb. Therefore, this Corpus still requires more elaborated part of speech tagging when compared to the parsing algorithm in COCA Corpus [2], which allows automated search of clusters with reference to part of speech delimitation, as can be seen in Figure 14. Taking this issue into account, the further investigation in EEBO refers to the contextual analysis based on syntactic and semantic criteria. Furthermore, in case of just functioning as an adverb, I specifically consider its meaning in the text to check whether it acquires a particularizer or exclusive function characteristic of this adverb in PDE.
To manually differentiate adjectival and adverbial usage of just and to avoid ambiguity of data interpretation a special attention is paid to syntactic regularities at every time-frame of under analysis. Taking into account that the adverb entered the English language in the XIV century [22], some specific features of clause structure can be summarized as follows:
1) The language of XIV-XVII century is known as the period of WO normalization with verbmedial turning into a canonical pattern [46]. 2) Syntax greater rigidity facilitates the consistent decline of V2 in the English language of XV century [47], as well as, modifications in VP and DP structures [48; 49]. obligatory used as a head [53]. 7) Adjectives frequently require either a dummy head or a pronominal one. [54]. If the DP is compiled of a set of adjectives, the second adjectival element can oftentimes follow the noun [34], as in a good man and iust. Once it's a case a dummy one can be observed or omitted from the surface structure of the sentence. Therefore, a special attention is paid to DP structure when considering just in its adjectival and adverbial meanings in cases when it occupies the position after the noun, given that it can function as an adjective in this period (12).
Another line for differentiation for adverb just presupposes its contextual analysis as a focusing adverb and other types such as manner and time [40], which is to be done manually. While functioning as a focusing adverb just may be used as an exclusive or a particularizer, hence, the next step of the analysis suggests the distinction of specifically these two types based on alternative semantics methodology, viz. Question Under Discussion (QUD) method [35; 55]. It should be noted that the common feature for focusing adverbs lies in their establishing relations between the Focus value and the set of alternatives [56, p. 340]. Therefore, the differentiation of focusing adverbs is made possible with reference to their discourse function, presupposition and descriptive content [44].
Adverbs of an exclusive type are aimed at providing a comment on the current question (CQ) that weakens a salient or natural expectation. Hence, taking into account the topicality scale, the antecedent should be weaker than the presupposed answer to the CQ. The presupposition for exclusives implies that they express one of the most likely true alternatives to the current question, which is "at least" as strong as its antecedent on condition that the last is the minimally expected true answer. This schematically can be displayed as follows: MIN(𝜋) (where 𝜋 is an antecedent) [57]. The descriptive value conveys the most likely true alternative for the CQ, which is "at most" as strong as its antecedent and can be represented as MAX(𝜋) [58]. Based on the abovementioned, the operators for upper and lower scale can be defined as in ( 13)-( 14 where ≥ 𝜎 is a pragmatically given pre-order on the propositions that constitute all the potential answers to the CQ.
Given that just may function as an exclusive adverb, its semantic meaning is exemplified in ( 15)-( 16).
The traditional content of just in (16) and its usage implies a scale over the probable answers to the CQ. Such a scale is reflexive and transitive and should not necessarily be antisymmetric. The presupposition demands from the pre-adjacent to just to be the minimum true answer to the current question or in other words to be the least likely true answer expected. The descriptive content suggests that the maximum true answer to the CQ should be no-stronger than the pre-adjacent. Therefore, the latter is not presupposed in advance, however, the QUD rules put forward in Beaver&Clark [57] ensure the presence of true alternative answers to CQ, as well as, the truth of the proposition (ОА).
The discourse function of particularizers confines in expressing identity and specificity. According to E. König [14] and A.-M. De Cesare [59, p. 65] they are used emphatically with their main aim to establish identical relations between arguments in proposition, which are contextually given though do not exclude other possibilities. Based on their presupposition the use of particularizers presupposes that the assertion is related to the greater part of Focus [25, p. 158; 60], with a particularizer emphasizing the validity of the sense but not its own validity compared to other non-valid values [61, p. 158]. In other words, it focuses on the pragmatic implicature of stating that something is true by nature, so two sentence elements can be evaluated as equal. [43, p. 348]. So, as a particularizer just can be paraphrased as 'nothing but' or 'X and only X' [48]. Thus, the speaker can presuppose the possible alternatives, contrastive meanings, however, when the sentence contains a particularizer adverb this sentence lacks contextual prompts due to the general unnecessity in that. E.g. 17. You look just like your sister.
ME Corpus data show that the polysemous form of just entered ME in ca. 1300s functioning as an adjective (19), a noun (the ancient Olympic games; A series of single combats) (20) or a verb (21), as well as, later as an adverb (22).
Analysis of ME Corpus data proves that out of 1022 matches in 63 records the form iust is most frequently used as an adjective while the nominal form appears to the least frequent one (See Table 3)
Table 3 demonstrates that iust(e) as an adverb despite emerging within the same time frames as its counterparts, viz. Adj, N, V, goes through the process of its early standing, which is confirmed by the low frequency figures in ME Corpus (0.69%). The dominance of adjectival usage can be observed with other spelling variant just(e) represented by 268 matches in ME Corpus. However, there is a notable difference in part of speech allotment ratio for this particular spelling form (See Table 4).
Allegedly, the presence of two forms is attributed to dialectal differentiations and lack of graphic standard. Yet, Corpus studies indicate that two different spelling variants may coexist not only in one text separately, but are even traced in the same sentence. E.g.
Comparing the data from Tables 3-4, the assumption is made that based on frequency figures the form just but not iust is more prone to categorial shifts and develops into the dominant adverbial spelling variant in further centuries of the English language evolution. The analysis of the two forms in EEBO presented in According to the OED [61], as well as, the tokens retrieved from ME Corpus adverbial meaning of just arises in 1417 presumably from the adjective juste, which has two primary senses: the first meaning of just refers to the persons, the heart, living morally upright, righteous (24); while the second major sense of just refers to the equitable or fair, fitting proper, conforming to the rules (25)-( 26). Apparently, both major senses of adjective just came to English at the same time, which is indicated by ME Corpus entries of 1384.
To speculate on the transformation of adjective just into an adverb, a special reference should be made to some examples from Corpus of Middle English Prose and Verse, where the form under analysis gets a double reading due to the post-modifying placement of the adjective and only the contextual analysis and knowledge of clause structure specifics helps differentiate adverbial and adjectival senses. E.g.
The occurrence of such ambiguous examples may testify to the verity of the hypothesis that adverbial meaning originates from the adjectival on account of dual interpretation of the clauses. As has been hypothesized in the paper, the early stages of such transformations involve metaphorical abstractions [23, p. 16]. This hypothesis finds evidence in Erman [62, p. 99], who claims that just (exactly) has been derived from the adjective with the meaning of fair, correct, precise. Overall, 20 entries of adverbial just elicited from ME Corpus indicate that the dominant particularizer meaning is associated with the notion of exactness. Hence, the major senses are distributed as follows: a) in an exact or accurate manner; so as to correspond exactly; with precision; accurately; punctually; correctly (11 Based on these figures, the preliminary conclusion runs as follows: the analysis of data from Corpus of Middle English Prose and Verse shows that just functions as the adverb only in 20 instances out of 1284 matches in total, while other forms of iust(juste) are distributed among adjectives, nouns and verbs. Therefore, the standing of just as an adverb starts in ME, undergoing further semantic and grammatical specialization in Early Modern English acquiring restrictive focusing adverbial function [63]. Hence, just overcomes extension moving to an open-class category.
Early Modern English findings from EEBO are initially analyzed in terms of frequency while rendering the adverbial meaning by two forms just and iust with reference to every decade up to 1650s.
The data indicate that just is typically associated with an adjective in the years of 1480s-1500s and its functioning as an adverb is limited to 2 instances where it is used in the meaning of justly (adverb of manner), therefore the data for this period hardly turn informative. E.g. The analysis of tokens with just in the years from 1510s to 1550s points to their gradual increase in usage per million words reaching the figure of 120.92 in 1550s (particularly for the spelling variant iust) (See Fig. 16). The ratio for just functioning as an adverb rises to 4.38% also demonstrating adverbial meaning extension represented in Table 6. Table 6 testifies to development of the polysemantic sense of just as an adverb with its dominant particularizer function (exactly, precisely). The meaning of amount and time rendered by the adverb can also be interpreted as right (exactly) characteristic of a focusing particularizer. Significant for this time frame is the rise of meaning justly traced with the spelling variant under study, since Middle English records have a specific form justli (jostle, justle) first entering the language in 1384 simultaneously with adjective just, which served the foundation for adverb justly formation [44; 61]. Therefore, the adverb demonstrates further meaning extension untypical for the previous English periods.
The next time frame, viz. 1550s up to 1600s, shows the significant decline in the meaning of exactly (manner), whilst the number of examples with just emphasizing the amount and time doubles. Moreover, a new function arises, i.e., modifying degree and comparison.
The distribution for 1600s-1650s (Table 8) shows the abrupt drop of the sense justly, whilst the particularizer meaning becomes dominant indicating mostly location or manner. Another interesting feature for this period is the emergence of just in its exclusive function, which can be replaced by a synonymous construction with only (34).
Adverbial meaning of JUST in 1600s-1650s Therefore, grammaticalization cline for just appears as hypothesized in the Introduction, with exclusive sense of just arising in 1620s. Firstly, adverbial specification is limited to highlighting something exact, then just goes through the process of meaning extension turning additionally into the adverb manner and time, which allows new senses to enter the domain of just in EModE. Moreover, as the examples demonstrate, adverbial sense of just significantly prevails after 1600s with the adjectival usage being still dominant. Thus, the process of just grammaticalization as a closed class category is still on its way in EModE, however, after 1620s it starts functioning exclusively as a restrictive adverb. Characteristic of these patterns is their occurrence in poetic records, hence, the sentence word order may have been affected by metrical requirements, since the information structural analysis shows no regularities in word order differentiations, viz. the element just is associated with in OSv+just→V(X) pattern is tagged as new information and informational Focus, while in S just→O(X)vV pattern the component following just conveys situationally accessible information and identificational Focus.
A greater amount of data on information structure peculiarities is available for Early Modern English period, where the figures are analyzed within such time frames: 1500s-1550s, 1550s-1600s, 1600s-1650s. The analysis of word order for every timespan shows that mostly just is represented in the patterns where it modifies XPs that follow the verb. This regularity is already dominant for 1500s-1550s. The WO patterns and information types as well as Foci variations are illustrated in Fig. 17 The years 1550s-1600s testify to the adverb greater positional variability especially for just in postmodifying placement affecting the general word order arrangement. In all these cases the information marked by just is represented as new referring to emphatic Focus. Significantly, the number of instances with post-modifying just triples in this period (See Fig. 18), which is not the distinctive feature of the next 50-year time frame, since the share for a rigid SVO order rises to 84% with just in a pre-modifying position following either the verb or the object (See Fig. 19). The time frame of 1660s-1650s displays a slight reduction of the ratio of marking new information and informational Focus (up to 38%) in the SVO pattern with the redistribution in favor of given information and identificational Focus (18.43%). This fact is evidenced as the crystallization of the particularizer function of just. Besides this, the ratio of clauses where the element marked by just renders emphatic Focus significantly arises for both 1550s-1600s and 1600s-1650s time frames with just predominantly occupying the first position in the sentence causing inverted WO.
Corpus methodology investigation suggested provides new insights in focusing adverb just analysis relying on automated, semi-automated and manual procedures to study the adverb graphic representation in ME Corpus and EEBO. Quantitative and qualitative data allow assuming that adverb just firstly registered in 1417 originates from the adjective due to ambiguous reading in the records prior to this date. Yet, the scarcity of the examples for this time frame fails to provide significant quantitative data on adverb grammaticalization. Therefore, it is assumed that it is not until 1500s that just goes through the grammaticalization process extending its original ME sense with the significant rise of adverbial component in 1600s-1650s, which is especially pronounced after the year of 1620. Thus, its initial meaning of exactly or precisely is enriched to render nuances of manner, amount, time, location, degree or comparison with further specialization on particularizer functions. Therefore, by 1650 it can already modify NPs, PPs, APs and VPs functioning predominantly as a focusing particularizer adverb, while its exclusive sense typical of PDE only becomes evident after 1600s. The quantitative data suggest that the process of just turning into a closed class category is still on its way in EModE, since the other non-focusing adverbial meanings are evident in EEBO Corpus till 1620.
The automated and semi-automated Corpus analysis by means of coreference-resolution annotation tool allows investigating WO and IS correlation with reference to focusing just, which major function by 1650s becomes marking the identificational Focus and situationally accessible or given information amounting to ca. 46% in total, which is typical of adverbial particularizer. This fact provides further evidence for the adverb grammaticalization. The role of the adverb in information structure highlighting is particularly prominent when it is adjacent to sentence element that represents emphatic Focus and either new or given information, since just pre-modifying position significantly affects WO variations causing object fronting in the clause.