The treatment of the semantics of just by Colins Cobuild Advanced Dictionary of American English ¹ is a typical one. It distinguishes between an adjectival just and an adverbial just, without explicating their interrelation, and it lists 16 meaning items all in all, again without clearly defining the criteria used in the semantic differentiation. In fact, the differentiation is so vague that one finds overlaps between meaning items or even within the same meaning items, and unclear interrelations between some fixed expressions, when otherwise necessary contextual information is unavailable. For example, the meaning item 2 reads like “If you say that you are just doing something, you mean that you are doing it now and will finish it very soon. If you say that you are just about to do something, or just going to do it, you mean that you will do it very soon”. In addition, while the meaning items 3, 5, 6, 10 and 14 are all concerned with the emphatic function of just, no elaboration is made on the condition and manner of the performance of the function. Likewise, both the meaning items 11 and 12 are concerned with the usage of just a minute/ moment, both the meaning items 15 and 16 are concerned with the usage of just about, which is somehow tangled with the meaning item 2, and further, the elaborations in the meaning items 4, 7 and 8 shun the senses of just in themselves, only making reference to some possible interpretations of just in specific contexts. For instance, the meaning item 7 goes like “You can use just in front of a verb to indicate that the result of something is unfortunate or undesirable and is likely to make the situation worse rather than better”.

The lexicographical semantic description of just, therefore, fails to offer a definitive account of its semantic structure and features, with a confusion of its semantic and pragmatic qualities such that most of the meaning items are not mutually exclusive. The loose ends pose practical difficulties and challenges to language users when faced with uses of just in various contexts, for example, being unable to interpret or explicate the ambiguity in the following case:

3) He just threw a ball to me. ²

The first meaning of just in the above example is associated with its adverbial function, stating that the throwing of the ball occurred a very short time before the sentence was produced. The second meaning is adverbial too, but it expresses an emphatic function, indicating that throwing the ball was the only action he performed. The intended meaning by the producer can be determined only through context.

Research findings now available, such as Cohen (1969) , Quirk et al. (1972, 1985) , Lee (1987, 1991) , Kishner and Gibbs (1996) , Lindemann and Mauranen (2001) , Grant (2011) , Duffley and Larrivée (2012) , Molina and Romano (2012) , and Beeching (2016) , among others, virtually adopt or support the view of polysemy of just ³ , arguing for its syntactic and semantic variability, with major differences in terms of the types and quantity of the senses of just. In summary, the following six primary ones have been identified in the existing literature ⁴ , mainly in the works of Lee (1987) and Kishner and Gibbs (1996) .

The depreciatory sense of just means that it is used to downplay or downtone certain events, acts, or situations, i.e., signaling they are minimally significant ( Lee, 1987 ). The evaluation serves as an indication of the speaker’s attitude or way of treatment, often in comparison with a certain criterion related to entities of the same type, but often not mentioned explicitly ( Gibbs & Matlock, 2001 ). That is, as analyzed in this study, a reference object (R) is negated while a focus object (F) is affirmed. Generally speaking, the values or degrees that R and F correspond to, respectively, are known to both parties involved in communication, which are often presupposed and easily activated with some modality. Roughly, just in this sense equates to merely or simply (Cf. (4)). Further, the contrast between R and F makes up a continuum, except that sometimes the contrast is not explicit enough (e.g. 5; Cf. Lee, 1987, 1991 ; Kishner & Gibbs, 1996 ):

4) It’s not serious—just a nuisance… (That one’s just a cyst, not a lump yet).

5) Just close your eyes.

The restrictive sense of just ( Lee, 1987 ; Wierzbicka, 1991 ) means that it is employed to restrict things, acts, or events in certain situations involved. Consider the following examples (Cf. Lee, 1987 ):

6) I just noticed it when I am in the kitchen.

In the utterance above, just does not semantically orientate or govern the subordinate clause, and the moment of “noticing it” is only restricted to the temporal point designated by the when-clause, while other possible temporal points are excluded. That is to say, the proposition of “P just when Q” and the implicated proposition of “P not just when Q” form a contrast. For that reason, the following sentence is unacceptable for its semantic self-contradiction (Cf. Lee, 1987 ):

7) *I just notice it when I’m in the kitchen, and I also notice it when I’m in the study.

Contrary to the depreciatory sense of just, the restrictive sense is functional in expressing conceptual or propositional meaning and thus is roughly the equivalent of only ( Wiegand, 2016 ).

The specificatory or marginal sense of just means that it can be used to specify the concrete spatial-temporal positions of or relation between two adjacent events or physical entities, i.e., marginally close to the explicit boundary or periphery of a process or something. As is well known, spatial entities can metaphorically refer to temporal ones, meaning that F is maximally close to R while one boundary of F is unclear and the other boundary becomes clear because of R. Examples include (1) and the following (8) (Cf. Lee, 1987 ):

8) It gets itchy just under the eyebrows.

Typically, when used in the specificatory or marginal sense, just is followed by PP or adverbs of place, and it is often interpreted as equivalent to immediately, right, barely, marginally, or fractionally (also Cf. Lee, 1987, 1991 ; Kishner & Gibbs, 1996 ). Sometimes such an interpretation is possible in certain tensed aspectual expressions. For instance:

9) He’d/He’s has just come out of hospital ⁵ .

The fourth type of sense of just is an emphatic one, which means that one can use just to highlight a specific part of a related semantic range so that a related concept stands out or becomes prominent ( Cf. Lee, 1991: p. 57 ). Just used in this sense is often seen modifying or in collocation with expressions indicating strong emotions such as so, infuriate, amaze, dazzle, terrorize, atrocious, at all, and so on (ibid), or in certain contexts associated with negation. Roughly, just used in an emphatic sense is equal to really, as indicated in (10):

10) (It’s just so hard to keep his balance.) He is just amazing/that’s just impossible!

The fifth type of sense of just is a comparative one, often used in comparative constructions and frequently seen before as or like, performing a modifying function. The comparative just is semantically equivalent to in the same way. Consider the following two examples:

11) I love chocolate, just as you love pizza.

12) Susan is just like her sister.

The sixth type of sense of just identified is termed as “exact”, meaning that the adverb is used to focus or lay stress on what introduces an embedded clause or the referent of wh-words or demonstrative pronouns such as that and these. In this case, just can be paraphrased as exactly. See the following example (13):

13) That’s just what the doctor ordered.

A major problem with the semantic categories identified above is that they often permeate each other and overlap in such a way that the categorical peripheries are not clear-cut, and, as a result, a large quantity of cross-categorical uses are found in our data (also Cf. Lee, 1987, 1991 ; Kishner & Gibbs, 1996 ). Due to the syntactic-semantic variability, the interpretation of just is found to be considerably dependent on contexts, especially on the interactions of it with other components that co-occur in the same utterance. For that reason, both Lee (1987, 1991) and Kishner and Gibbs (1996) admit that such a problem disproves or undermines the validity of the polysemy account of just. Therefore, as it is argued here, a cross-categorically oriented analysis can be more insightful. Different cross-categorical uses are discussed below.

Brown & Levinson (1978: pp. 181-182) argue that just in (2) depreciates the significance of the content of the related proposition by weakening it to an extreme. Similarly, Lee (1987, 1991) holds that the primary semantic difference between the two uses of just in (1) and (2) is that just used in the former takes part in the expression of the propositional content of the sentence, while just used in the latter does not. Instead, it only functions as an orientation of what is expressed by the proposition to the role and attitude of the participant involved in communication. Therefore, the former performs the ideational function and the latter performs the interpersonal function in Halliday’s terms. As such, the objective description ⁶ and subjective expression of one’s attitude or stance brought about by the use of just are often intertwined and thus difficult to distinguish, and, indeed, speech acts in social interactions are often subjective and the modal use of just can make an answer sound more polite, which helps to save or maintain a hearer’s face. Consider the following mini-talk (Cf. Lee, 1991):

14) Doctor: Was there anything else?

Patient: Oh, I was just going to talk about my husband.

As indicated in the above example, the doctor wants to put an end to the talk, but the patient is surprised and says she is “on the edge” or at the moment of talking about her husband, which on the one hand indicates that just has a specificatory/marginal sense, but on the other hand can be interpreted as meaning that talking about her husband is not important at all (to the doctor and/or to herself). If just is understood in this depreciatory sense, the patient may imply that she is willing to cooperate and terminate their talk, or she is declining the doctor’s request and suggests that she would rather go on with her talk about her husband. It is clear that both the depreciatory sense and the specificatory/marginal sense are associated with “restriction”, but just is often taken to carry the depreciatory sense when the speaker’s attitude or stance is alluded to.

The contrast between these two senses involves a striking difference in a speaker’s attitude and emotion when using just. Based on such a feature, Quirk et al. (1985: p. 584) argue for the polysemy treatment of just, with the only difference lying in such paralinguistic features as tones in the expression, and, sometimes, the semantic features of the co-occurring components. Lee (1991) also points out that the emphatic use of just is frequently accompanied by high-pitched tones, intense sound, fast and brief breath, as well as other expressions indicating strong emotions such as at all, not a single, not a damn, etc., and corresponding body language such as facial expressions and shrugging; if the tone is neutralized and mild, just is often taken as depreciatory. Since it is obviously a matter of degree, when the aforesaid features are missing, determining the meaning of just usually becomes a tricky task. See the following example:

15) He just didn’t listen/wasn’t paying attention.

There are quite a lot of cases where the semantic features of the co-occurring components of just may affect its interpretation, for example, when negating epistemic modals or root modals, or when expressions that strengthen or weaken emotions in such contexts are involved. Consider the following three examples (Cf. Lee, 1991):

16) He just may not/won’t/can’t have seen us.

17) He just won’t lift a finger.

18) He just can’t do it, unfortunately/I’m afraid.

According to Lee (1991) , the speculative meaning in (16) becomes stronger and stronger one by one: when just is collocated with may, which has the least speculative force, it most probably carries the depreciatory sense; as won’t has a slightly stronger speculative force in comparison with may, and it already shows an affirmative color ⁷ , just used as a modifier of it is more likely to perform an emphatic or depreciatory function; when just collocates with can’t, which has the strongest speculative force, it most probably carries the so-called emphatic sense, especially when the speaker is refuting the assertion of its talking partner ⁸ . The underlined part in (17) indicates strong emotion, so just in it has an emphatic meaning; however, the underlined part in (18) weakens the expression of the speaker’s emotion, so the function of just in it is rather vague ⁹ . Therefore, it seems that the depreciatory sense and the emphatic sense of just form a continuum in some cases, if not all, and its interpretation in specific contexts shows a gradience because it is conditioned by various contextual factors. Consequently, the semantic indeterminacy of just in language use is quite common.

A precondition for the determination of the specificatory sense of just is that one end of the related semantic range is open and therefore has a vague boundary, and the other end is immediately adjacent to R and therefore has an explicit boundary. Good examples are (1) and (8) given above. In (1), the temporal point immediately approaching the point of “midnight” (e.g., R) is clear (say, suppose it is 23:59 PM), and, in (8), the spatial point immediately adjacent to the physical boundary of the “eyebrow” is also explicit. In both cases, the notions of “midnight” and the “eyebrow” are definite with clear boundaries so that just indicates a point belonging to the marginal section. Therefore, any other points that do not belong to such a marginal section are excluded in a truth conditional sense in the interpretation of just. For instance, in (1), the temporal point of 23:58 PM or other points than 23:59 PM do not apply. However, in some cases, the notion of space and sections of space can form a continuum, and some spatial points on the continuum can apply in the interpretation of just in a truth conditional sense, with differences only in degree. Take the following example (19) (Cf. Lee, 1991):

19) The shop is just near the bank. Jo was standing just at the corner.

In the two uses of just in (19), in accordance with Lee (1991) , the physical boundaries of the two entities, i.e., “the bank” and “the corner”, are clear and definite, but the adjacency to them makes up a continuum, and the notion has a feature of gradience. That is to say, the more F is adjacent to R, the more reasonable and acceptable the sentence becomes. Thus, all the points that are reasonably at the gradience are applicable or valid as candidates in the interpretation of just in the case in the truth-conditional sense. Such a gradual process implied by the gradience proves that just has an intensification sense ( Cf. Lee, 1991 ) ¹⁰ , where the R end has a clear boundary. In contrast, the two ends of F and R, when the specificatory/marginal sense of just is involved, have clear boundaries, and the emphatic sense of just implies intensification, except that the two ends in this case do not have clear boundaries, which is observable in the uses of amazing and impossible in (10). Similar to the emphatic sense, the intensification sense of just highlights a prominent part of the features related in the continuum, but, in spite of the distinction made above in terms of bounding feature, cross-categorical uses of just are still possible when a speaker conceptualizes a situation before he or she expresses himself or herself. Take the following example ( Cf. Lee, 1987, 1991 ):

20) I was just about to leave.

Lee (1991) argues that the notion of “periphery” exists independently in the notion of “being about to leave”. Then, when it is taken as a reference point with a clear and definite boundary, just carries a static specificatory/marginal sense; however, if the situation is taken as a dynamic process in which “I” was approaching the temporal point of “leaving”, then just carries the intensification sense, but it carries the emphatic sense if prominence is given to the state of “being about to leave”. Therefore, it is obvious that the differences in the interpretation of just in the same sentence are the results of different construals (Cf. Langacker, 1987: pp. 63, 117). Additionally, the way of construal is subject to communicative situations involved and one’s encyclopedic knowledge. Take the following examples (Cf. Lee, 1991):

21) The ball landed just outside the line. vs The car stopped just outside the bank.

Construals in different epistemic domains (Cf. Langacker, 1987 : pp. 63, 117) will lead one to different cognitions of the relative spatial positions of “the ball” and “the car” to “the line” and “the bank” respectively when one consults his or her encyclopedic knowledge about the specific entities. For that reason, the two justs as used in (21) demonstrate a contrast between the specificatory/marginal sense and the intensification sense, i.e., in the former sentence, just shows a stronger specificatory/marginal sense, while in the latter, the intensification sense is more remarkable.

The restrictive sense of just concerns an objective expression of propositional content itself, rather than the subjective expression of a speaker’s attitude or evaluation typical of the depreciatory sense. The example in (6) demonstrates that point, and consider another one here (Cf. Lee, 1991):

22) Just in one heel it lifted, but not its back in both.

It is noteworthy that the restrictive sense of just involves entailment, while the depreciatory sense is associated with contrast, in which entailment is unavailable or very hard to determine. For instance, the sentence in (6) entails the one in (23) (Cf. Lee, 1987 ):

23) I notice it when I am in the kitchen.

Let’s take a look at the interactions of just with negation and the related scopes in the following sentences before we proceed to offer an explanation of the difference and contrast implied in the aforementioned examples (also Cf. Lee, 1991 ):

24) I just didn’t speak.

25) I didn’t just speak.

One can see clearly that in (24), the negation falls within the scope of just, in which case just is often taken to carry the depreciatory sense. However, in (25), by contrast, just is often understood to be restrictive in sense. Lee (1991) argues that it is a result of pragmatic influence, in that the situation is concerned with a contrast between R and F. Specifically, when F is governed by just, it calls for an understanding of the adverb as depreciatory or restrictive in sense, but negating a proposition, as a way of expression, can be employed to modulate the tone and weaken the depreciatory force because it is often used to refute a prior assertion or entailment made and makes itself an appropriate means of modalization. Nonetheless, when making a choice from many alternatives is involved, one often gives prominence to just in his or her expression of the restrictive sense, but ultimately its interpretation is dependent on various contextual factors rather than on the linguistic expression itself. The following example suffices to illustrate this point:

26) (A says that B didn’t listen or speak, and B answers) No, that’s untrue. I did listen carefully. I just didn’t speak.

This way of negation achieves the same result as in the negation of “just VP” in (25), which is the restricted item. It is also not likely a case of negation of “just VP” in the depreciatory sense. Whether just is taken as carrying the depreciatory sense or the restrictive sense is ultimately determined by the interaction of just with negation and the relation of F to R. Importantly, the ways of construal made by the two parties in communication greatly affect the interpretation of just. For example, if an explicit R is missing, their interpretations are often different.

In summary, it is argued that the polysemy-based semantic account is essentially an analysis and description based on pragmatic functions in contexts such that the senses identified are closely interwoven and permeate each other. As a result, since the senses are not exclusive enough among themselves, the interpretation of just in utterances ultimately relies on the complex interactions among contextual factors. In fact, both Wierzbicka (1991: p. 350) and Aijmer (2005a) point out that Lee’s (1987) classification of the senses is too elaborate and detailed to explicitly define the semantic nature of just as well as the internal differences and relations among the senses. Especially, the polysemy account fails to demarcate the semantic and pragmatic similarities and dissimilarities between just and its synonyms or semantic equivalents such as merely, only, and simply, etc. Lee (1987, 1991) himself is well aware of the limitations of his polysemy account and admits that a semantic account with more descriptive and explanative power and higher operability is called for.

Traugott (1989) argues that one needs to examine the lexical semantic development so as to ascertain the evolutionary path and the general trend of the change if they intend to achieve the internal reconstruction of the lexical meaning and account for polysemy. In accordance with the findings of Cohen (1969) , Traugott (1988, 1990) , Nevalainen (1991) , Preston (1993a, 1993b) , Molina & Romano (2012) , and Beeching (2017) , just originated from a Latin adjective meaning “righteous, fair, legitimate, equitable” and was borrowed into English via French in the later period of the 14^th century at the latest. The archaic meaning is in use even today. No later than the year 1391 did the sense of “precise” appear, and very naturally the corresponding adverbial use of just came out while the semantic extension was in progress, i.e., “precisely, exactly”, along with the use of just as an adverbial of manner, meaning “justly, honourably”, which as a critical usage helps change just into a scalar particle ultimately. It should be noted that just used in this sense has a synonym “right” which can mean “clarification”, which in turn well matches “precisely”. As a scalar particle, just is used to give prominence to F, and meanwhile it associates as a particularizer with the values of other choices at the same related scale ( Cf. Traugott, 1988: p. 131 ). It is generally held that the sense of “precise(ness)” is a critical point along the semantic evolution path from which other senses or interpretations of just extend in various directions.

According to Nevalainen (1991: pp. 151-155) and Preston (1993a, 1993b) , at the beginning of the 15^th century, just developed the specificatory/marginal sense used for expressing positional precision in measurement, which derived from the sense of “exactly, precisely” and became prevalent in use in Early Modern English (EME). Till the later period of the 16^th century, the specificatory/marginal sense extended into the temporal domain. Specifically, just was used to perform the absolute temporal function in the middle of the 17^th century, i.e., it modified F at the moment of speaking (e.g., She left just then); and then later it could modify a nearest future temporal point (e.g., She’s just leaving), and even later it could be used to modify a nearest past temporal point (e.g., She just left). In such a way, just experienced a change from directly modifying a temporal NP to modifying a VP, thereby becoming closely associated with tense and aspect. It is noteworthy that such a semantic extension of the specificatory/marginal sense from the spatial domain to the temporal domain and the subsequent evolution related to temporal references are all reflections of human construals and subjective treatment, representing a subjectified process of making objective expressions of the real world and subjective expressions of the unreal world. For that reason, Traugott (1988: p. 134) argues that just is gradually merged with the speaker’s subjectivity in his or her expression of propositional content and thus ultimately is no longer used to indicate precision of measurement. Understandably, the uses of just in the depreciatory sense and other uses indicating temporal points in the future or the past are no later than its use as a scalar particle.

The investigations made by Nevalainen and Preston reveal that the restrictive sense of just appeared in the late period of the 17^th century when both the older and later usages were available side by side. The exclusive sense of just (i.e., exclusively) did not emerge until the end of the 17^th century. Meanwhile, uses of just based on its pragmatic functions and the subjectified expressions became even more prevalent and more and more mature until, in the early period and later period of the 18^th century, increasingly subjective use and downtoning use of just came into existence respectively, the former being equivalent to really/simply and the latter being synonymous with merely. It is important to note that, while other senses or functions of just kept coming out sporadically in history, such as the sense of “indeed” that appeared in the 16^th century and the intensification sense in the 18^th century became even stronger so that just came to perform an emphatic function, the general trend is that in English in the later period of the 19^th century and Modern English, just gradually lost its semantic content and ultimately came to be used as a pragmatic marker functioning as a mitigator or minimizer to achieve politeness. To summarize the major findings of the aforementioned studies, the semantic evolution path of just can be shown in the following chart ( Cf. Preston, 1993a, 1993b ):

From such a chart ( Figure 1 ), one can easily spot that the “preciseness” sense of just is a key point and condition in its semantic evolution because it is from such a point that the adverbial use of the word, i.e. “precisely/exactly”, is possible, from which, in turn, semantic extension in various directions occurs. Specifically, along with the semantic change within the social-physical domain, usages in epistemic domain develop in which the subjectified extension proceeds along two routes, i.e. showing prominence and demonstrating depreciation. Ultimately, the former usage develops into an adverb indicating emphatic function, while the latter usage changes into a pragmatic marker functioning in the speech-act domain as a diminisher and mitigator for politeness purposes. Roughly, the semantic structure of just appears to be central-radial, while the semantic evolution path shows apparent linearity. Such a changing process matches the trend description made by Preston (1993a, 1993b) and conforms to the pattern of lexical semantic change proposed by Traug ott (1989) , which follows a universal functional order of “Propositional > Textual > Expressive”. Traugott points out that lexical meaning evolves in a linear way in three conceptual domains, and that is, a word, such as because, is used to make an objective description of the world, as a propositional expression, in the social-physical domain, and then it is further used in epistemic domain within a text to indicate a speaker’s cognition or evaluation of the propositional content, and then further it is used to show the speaker’s evaluation and treatment of his or her own speech act in the speech-act domain. In addition, the subjectified expression is effected through making pragmatic inference, which is conditioned by communicative information and the principle of intensification of pragmatic relevance ¹¹ .

Lee (1991) speculates that, probably, just is relatively semantically homogeneous and therefore should be treated in a unified way. Similarly, Aijmer (2002: p. 173) treats just as a metalinguistic operator which provides guidance for interpreting certain linguistic signs. In the same way, Wierzbicka (1991: p. 350) argues for using a highly abstract generalized form for its semantic structure description ¹² . Traugott (1989) points out that polysemy has fuzzy sets as its structural feature and dynamic flexibility based on prototypical sense, based on which, as a semantic core, various new senses are derived so as to flexibly meet the demands of communicative contexts. Importantly, Traugott (1988, 1989, 1990) , Nevalainen (1991) , Molina and Romano (2012) , and Duffley and Larrivée (2012) all refer to scalarity for analyzing the semantics of just, and they believe that just in collocation with many syntactic components, such as any, usually leads to scalar interpretations. Such a belief is in accordance with the function of modulation proposed by Lee (1987) . Based on those considerations, it is hypothesized in this study that just, though having quite a lot of different meanings derived and extending in various directions through making pragmatic inference ¹³ ever since the meaning of “precisely” came out, has only one core sense; that is, it prescribes a restriction to a feature of a certain scale or to the minimal value of a certain scalar system and functions as a scalar restrictor. The scalar feature involved or activated is conditioned by context, and the interpretations of the same core sense of just used in the three domains can differ greatly due to different pragmatic functions.

Most, if not all, cognitive linguists hold that semantic fuzziness is one of the fundamental features of language use. Based on their embodied cognition and through their cognitive processes, human beings are able to know the world and express it with language. Determined by reality, on the one hand, human cognition determines language use; on the other hand, as a tool for and result of cognition, language use reflects cognition. Understandably, therefore, as conditioned by their own organs and faculties, human beings’ cognition of the world is always fuzzy, and, therefore, their use of language for expressing the world can hardly be absolutely precise. For example, the same modifier “red” in “red apple” and “red hair” may in fact refer to different colors. The semantic underspecification of this sort ( cf. Blutner, 2002 ) is a universal feature across language uses. As is well known, language users are always making choices and maintaining balance between achieving communicative efficiency and maintaining economy of language in using language. Likewise, in using language and expressing themselves, they are always trying to keep a good balance between being precise enough and being fuzzy enough. The Gricean Cooperative Principle with the four Maxims, especially the Maxim of Quantity, is a reflection or illustration of this law. Significantly, both decoding and encoding of semantically precise and/or fuzzy expressions are primarily dependent upon the purpose and demands of communication, and means of making modification are necessary if language users want to increase the degree of precision in their expressions.

Gatewood (1983) argues that talking loosely is a universal feature of human communication. Sperber and Wilson (1986) point out that due to semantic fuzziness, literal talk, loose talk, and metaphoric expressions are identical in nature and in ways of interpretation, with differences only in degree of looseness. Additionally, the looseness of human talk is pragmatic and non-truth-conditional in essence, which implies that people’s utterances do not necessarily follow the social convention of being literally true, and utterance interpretation cannot be absolutely accurate and true in every way to the propositional content of the utterances and the speaker’s intent, but only stays, at most, maximally close to them. Based on such an understanding, Lasersohn (1999) proposes that pragmatic slack or looseness ¹⁴ be used to account for the differences in people’s cooperation under specific contextual conditions. He further argues that both expression and interpretation are simply close to truth (value) to certain degrees, and the semantic differences in utterances often do not lie in truth conditions, but rather in how much pragmatic looseness they allow. That means how much they are allowed by contexts to be deviated from truth without being misinterpreted. To effect correct interpretation, people will make use of encyclopedic knowledge they possess and all sorts of real-time non-verbal communication factors in their mental process of interpreting, such as body language and tones, etc. Normally, both parties in communication can tolerate or ignore unimportant fuzzy information in their language use if pragmatic conditions are satisfied, especially discourse relevance. Take the following examples ( Cf. Lasersohn, 1999: pp. 522-523 ):

27) Mary arrived at (exactly) three o’clock.

28) (All)/The townspeople are asleep.

Due to the semantic contribution of exactly, the relevant meanings both encoded and decoded by utterance (27) can be any other temporal point than “three o’clock sharp”, and the point is extremely close to the latter temporal point. The use of all in (28) performs a similar function and achieves similar effects. Utterance (28) without the use of all normally is interpreted as already containing the concept of all ¹⁵ ; however, whether or not “the whole of the townspeople” are asleep does not affect the communication in significant ways only if the utterance meets the contextual requirements. Whether what is encoded in or decoded from the utterance should be an absolute “whole without any exception” or not varies across contexts ( Cf. Lasersohn, 1999: p. 525 ). Lasersohn then points out that language users make use of certain linguistic means or forms to modulate or regulate the information transferred in their utterances so as to decrease or reduce pragmatic looseness. Such slack regulators as exactly and all indicate the amount of pragmatic looseness contained in the utterances. They are targeted at the so-called pragmatic “halos” and thus can make the expressions more precise. It is here argued that just is another slack regulator that falls within the category Lasersohn discusses.

Normally, there is a core sense lying at the center of the lexical semantic structure from which other related senses derive in different contexts. However, such a traditional definition is highly controversial in terms of the criterion used in sense differentiation, the nature of and the interrelations between the senses differentiated, the mechanisms of semantic derivation, etc. A notoriously difficult case in defining polysemy is related to over, a single word that has a complex lexical categorical status. Obviously, new ways of analyzing lexical semantics related to polysemy as traditionally defined are necessary ( Cf. Blutner, 2002 ; Nerlich & Clarke, 2003 ). In fact, quite a lot of so-called polysemous cases identified mainly under the influence of the thought of semantic compositionality ( Cf. Pelletier, 1994 ) are not genuinely polysemous when viewed from a Construction Grammar perspective. Consider the following cases ( Cf. Pustejovsky, 1993 ):

29) a) a fast car: (ambiguous) a car driven quickly/one that is inherently fast.

b) a fast typist: the person performs the act of typing quickly.

c) a fast waltz: the motion of the dance is quick.

d) a fast book: one that can be read in a short time.

e) a fast reader: one who reads quickly.

Pustejovsky (1993) distinguishes at least three senses of fast, arguing that the interpretation of the word is ultimately dependent on the context in which it is used:

30) a) to move quickly;

b) to perform some act quickly;

c) to do something that takes little time.

It is arguable that fast used in all the cases in (29) has only one prototypical sense, i.e., put simply, “(NP) moving at high speed”. Based on metonymical thinking and expression and frame semantics-related knowledge, one can use the “Art_def. + Adj. + NP” construction in a wide variety of ways. So, the three senses identified in (30) are actually different pragmatic interpretations of the same core sense in contexts. Specifically, the prototypical NP in the phrasal construction relates to an action or a process. And based on the prototypical use of the construction, one forms a mental schema in its mind along with a frame semantics system that relates to and matches the schema. In using the construction to express the same event semantics, a certain part involved in the related event catches one’s most attention and is given prominence in linguistic expression, encoded by the NP. Using any NP related in expression will activate one’s event semantics knowledge, and the degree of difficulty of the activation is largely dependent on how far the NP in question is from the prototypical NP in one’s mind and how familiar he or she is with such knowledge. Anyway, the activation of the same event semantics is metonymical in nature. When a non-prototypical NP is used, a hearer usually needs to resort to making invited pragmatic inference in interpreting the meaning of fast. In this sense, fast can be semantically underspecified when modifying certain types of NP in the construction ¹⁶ . The specific interpretations of a lexical item in a phrasal construction are results of constructional coercion, which does not lend support to the polysemous claim that different senses exist and affect the meaning construction in different collocates. Such a feature of constructional coercion is also seen in sentential constructions ¹⁷ ( Cf. Goldberg, 1995: p. 9 ). Nonetheless, Pustejovsky (1993) admits that the subtle differences between the senses identified basically are not independent of each other, but rather, they overlap and permeate each other and closely interact with each other. This viewpoint coincides with the spirit of Sperber and Wilson (1986) on loose talk. In light of what has been discovered and discussed so far, a unified account of the semantics of just is made.

Scalar structures represent human beings’ cognition and abstraction of degrees of qualities that things or entities possess, such as big vs small, far vs near, tall vs short, fat vs thin, heavy vs light, etc. They are diverse and various, and universal across language use as well, functioning as presupposition for language use and relating to both entailment and implicature of utterances. Lexical items and utterances in discourses are often associated with and/or trigger language users’ concepts and knowledge about certain scalars, thus forming or activating corresponding image schemas ( Cf. Fauconnier, 1975 ; Israel, 2001 ; Kennedy & McNally, 2005 ; Beavers, 2007 ; Katsos & Cummins, 2012 ). Additionally, both Lee (1987, 1991) and Duffley and Larrivée (2012) refer to interactions of scalar restriction between the senses of just, and Duffley and Larrivée even propose that just has a unified semantic representation of “goodness of fit”. Similarly, Coppock and Beaver (2014) treat exclusives like just as scalars and Fauconnier (1975) analyzes any as a pragmatic scalar trigger. It is argued in this study that just, like any in many ways, functions as a regulatory restrictor and regulates the scalar features or information amount related by restricting them to a minimal value range, thus reducing the pragmatic looseness. What is regulated varies according to what just modifies. Except for it being used as a discourse marker, as our corpus data reveal, just mainly modifies AP, NP, PP, VP, and clauses introduced by conjunctions, and it is not much conditioned in terms of tense or aspect. It is found in our BNC-based data analysis that, as a general rule, just holds a syntactic position that is almost always immediately to the left of the head it modifies ¹⁸ .

Take some gradable adjectives as examples ( see Kennedy & McNally (2005) for detailed discussion ). Quite a lot of semantically related adjectives are distributed on a scale according to their differences in grade, each pair of two of them forming a contrast, e.g., “He’s not funny, he’s hilarious”. Now I use an arrowed line section to illustrate the scalar modulation function of just on a scale or continuum of quality (cf. Lee, 1987 , 1991; also Laparle, 2016 ):

31) a) He is good.

b) He is just good.

c) No, he is not (just) good, he is fantastic.

Let’s suppose the value 0 means “mediocre”, and on its right, the one part of the scale as a whole is termed or conceptualized as “goodness”, and by contrast, the other part of the scale that stands to the left of the value 0 is “badness”. A group of semantically related adjectives are distributed along the line with two arrows in opposite directions, and they make up a scale like the next ( Cf. Laparle & Truswell, 2018 ; Laparle , 2016 : p. 20 ):

In (31), good in utterance a is open-ended rightward, and the concept of “goodness” has no boundary or upper limit, which means it includes various grades and causes semantic fuzziness and pragmatic looseness. However, in utterance b, the speaker uses just to restrict the related value of “goodness” to the minimal one on the scale, i.e., a specific mini section immediately adjacent to the point of 0, whose boundary overlaps with that of the adjacent concept of “great” (as indicated in Figure 2 , and note that the boundary between two adjacent concepts is usually vague), indicating the minimal value of the scalar concept of “goodness”. Narrowing the range of a concept like that is the so-called scalar restriction or regulation of pragmatic looseness, which helps maintain optimal relevance ( Cf. Sperber & Wilson, 1986 ). Importantly, utterance c indicates that on the part of the scale rightward, there are multiple other grades (represented by different adjectives) whose values make up contrasts with the specific value of “just good”, such as fabulous, amazing, stellar, perfect ¹⁹ , etc. The concept of “badness” to the right of the point of 0, as a reverse mirror image of the concept of “goodness”, can be treated likewise.

One’s knowledge on scales is activated by relevant language use in contexts. As discussed, the “exact”, “specificatory”, “restrictive”, and “exclusive” meanings or functions of just all derive from the meaning of “precisely” through one’s making invited pragmatic inferences. It is well understood that one naturally associates “precision” with “specification” and “exactness” of value on a scale, which again naturally connects to “restriction to a minimal value”, which in turn stimulates an association with “exclusion of other values on scale”. All the meanings are related to the objective description of things ( Cf. Molina & Romano, 2012 ; Traugott, 1988: p. 139 ) ²⁰ . Utterance (8) is an example that shows how just regulates a feature value on a spatial scale. In (8), the facial part that starts downward from the “eyebrow” as a reference point forms a range that functions as an open scale of spatial length, on which the “itchy” point is immediately next to the reference point on the same scale. Thus, the use of just places a restriction on a minimal value. The close associations and entailments between “exact”, “restrictive”, and “exclusive” meanings of just can be seen in (6) and (7). More detailed discussion on the “exclusive” meaning can be found in Coppock and Beaver (2014) and Wiegand (2018) ²¹ . It is noteworthy that the unexplanatory sense ²² of just identified by Wiegand again is in nature a kind of pragmatic-functional interpretation of the core sense of just in a specific context. It shows that one’s mental construction of a scale in communication is a highly subjective process, in that the speaker can use just to modify F only if he or she thinks that there is an R that is associated with F and meanwhile forms a contrast with it. Consider the following three examples ( Cf. Laparle, 2016: pp. 22-24 ):

32) A: I think Tom is just perfect.

B: Just perfect?

A: Okay, he is incredible, amazing, gorgeous, my one true love.

33) Tim is just witty.

34) Oh, Mary is just asleep. She isn’t dead or anything ²³ .

In the mini-talk in (32), speaker B thinks that A’s use of “perfect” is not appropriate enough for describing its comment and feeling on Tom, or summing up the good qualities of Tom. Utterance (33) involves ambiguity in that a possible interpretation is that the speaker may think Tom’s being witty as a quality is not enough according to a certain criterion or standard. Perhaps, he or she thinks that Tom should possess other important qualities. In (34), the speaker does not think any other more serious condition is involved on the scale of “degree of seriousness”, such as “dead”.

Temporal expressions are metaphorical expressions of spatial expressions ( Lakoff & Turner, 1989: pp. 99-100 ; Lakoff & Johnson, 1999: pp. 30, 139-140 ). The one-dimensionality of time connects the past, current moment of speaking, and future, which form a continuum or scale, which is closely related to one’s encyclopedic knowledge. Some modifiers, such as soon and recently, can be used to modify features related to temporal duration or point (Cf. Beavers, 2007 ). Similarly, just can be employed to regulate the temporal positioning or duration of events. Moens and Steedman (1988) argue that an event may be decomposed into three parts: preparatory process, consequent state, and culmination. And culmination is a punctual point of instantaneous change between the other two parts, indicating the completion of the core sub-event. Both preparatory process and consequent state consist of various sub-events that are all associated with culmination, and they as a whole are a part of a language user’s encyclopedic knowledge. Therefore, in virtue of metonymy, one’s speaking of any of the sub-events can activate its hearer’s knowledge about other related sub-events. The modification of VP by just follows one and single mechanism, with differences in the object of reference. For example, completing an MA thesis consists of a series of sub-events whose temporal continuum forms a scale, as shown below (Cf. (9); Laparle & Truswell, 2018 ; Laparle, 2016: p. 27 ) ( Figure 3 ):

35) a) Tim has just finished his paper.

b) Tim has/had just finished his paper.

c) Tim is just about to finish his paper.

d) Tim is just finishing his paper. (Then he will meet us at the bar.)

In utterance a of (35), R as a temporal point is the present moment of speaking somewhere on the scale to the right of the point X. When modified by just, the focused temporal point is regulated and shifted to somewhere to the left of and maximally adjacent to X, making prominent the most recent completion of the action. The same interpretation can be made of utterance b, the only difference being that R is a temporal point somewhere in the past. Similarly, R in utterance c is the present moment of speaking somewhere on the scale to the left of the point X. When modified by just, the focused temporal point is regulated and shifted to somewhere to the right of and maximally adjacent to X, making prominent the most recent beginning of the action. Similar to the case of utterance c, in utterance d, the durative temporal interval indicated by the progressive aspect is approaching the culmination point on the right very soon, but just as a modifier restricts it to the starting point of the interval.

Language use is naturally and necessarily subjective, and the use of just can easily affect subjectification of utterance making (Cf. (32) - (34)). Such an effect is achieved through integrating into expressions a speaker’s construal, attitude, evaluation, feeling, and emotion, etc. because uttering the word just is accompanied by phonological, facial, and other paralinguistic features. The critical condition to be satisfied for such an effect is the possibility of making an invited inference based on the restrictive sense of just. For example, the restrictive and exclusive functions naturally lead to the prominence of certain features in their extensions in different directions: through making an invited inference, one naturally connects feature prominence with intensification, laying emphasis, and depreciation (of others), the major pragmatic functions that all involve an increase in pitch and tone strengthening. In addition, restricting to a minimal value is easily associated with self-depreciation and mitigation, which are often accompanied by low pitch, slurring or swallowing syllables, or words not fully pronounced. Clearly, the matchup between such phonological phenomena and a speaker’s emotional attitude is a manifestation of linguistic iconicity. With the constant loss of its semantic content, the mitigating function of just ultimately evolves into a pragmatic marker that a speaker uses to evaluate his or her own speech act. When one restricts to a minimal value with the use of just, he or she is extending metaphorically the function to one that means “extremely trivial or unimportant”, indicating that the speech act itself is nothing important at all. In that way, as its semantic content has practically been lost and the tone is minimized, just sounds like a mouth filler and makes the speech sound more polite, especially when self-depreciation is made by the speaker ²⁴ . However, Wierzbicka (1991: pp. 350-351) argues that the term “depreciatory” is misleading because the evaluation is not always depreciatory: in some cases just can indicate intimacy, sweetness, comfort, self-defense, regret, or compliment, so on and so forth ( Cf. Traugott, 1988: p. 131 ). Consider the following utterance:

36) Let’s have dinner together—just the two of us.

Wierzbicka holds that the use of just in (36) shows a warm and sweet tone. Likewise, the semantic interpretations of subjectified uses of just are so subtle and complicated that they are largely context-dependent. On the one hand, just is associated with certain contextual backgrounds and scales, but, on the other hand, a subjective color can be attributed to such a use (Cf. (4)). Nevertheless, all the interpretations in contexts are ultimately based on its core sense: the restriction to the scalar minimal value is related to the concept of “smallness”, which in itself is an objective description, but how such “smallness” is valued, whether as good or bad, is largely a result of subjectification and social conventionalization ²⁵ . Therefore, parties involved in communicative interactions can make interpretations of the same utterance that are different from or even opposite to each other, rendering just functional in serving various pragmatic purposes. The unexplanatory sense of just mentioned earlier, for instance, is associated with a scale of “being able to offer more possible explanations”, and just is often used to indicate that a speaker can only offer a minimal amount of information, often with a strong emphatic tone. For example ( Laparle , 2016: pp. 36, 37 ):

37) a) The museum is just closing. No one knows why.

b) (Just be careful…) He is just dead. What else do you want me to tell you?

However, using just can also mean that a hearer is not allowed to make his or her own choice, or ask any further questions or make requests, which makes the utterance sound simple and crude. This is clearly perceivable in (37)b and imperatives like (5).

Although interpreting the core sense of just is a highly complex and flexible process with dynamics that are very much dependent on context, it can be relatively concrete and stable when just is used in specific syntactic constructions coupled with certain paralinguistic features. Kishner and Gibbs (1996) discover that, to a great extent, interpretations of just are conditioned by the syntactic environment in which it is used. Specifically, when modifying NP or AP, just usually has a depreciatory function, while it normally has “specificatory” or “exact” meaning when modifying interrogative or demonstrative words. And it often has “restrictive” or “specificatory” meaning when it modifies PP and VP. As noted earlier, the depreciatory and specificatory functions are associated with “restriction”, and just often carries a depreciatory function when a speaker’s emotional attitude is involved. However, just is frequently taken to be emphatic when it is used to modify a word that has a high scalar value along with strong subjective color (e.g., usually coupled with paralinguistic features), such as perfect, amazing, etc. On the contrary, it has a depreciatory function when the words modified have relatively smaller scalar values, such as good, mediocre, etc. And, when just is weakly stressed, it normally functions as a pragmatic marker. Consider the following utterances:

38) a) (Do you think the movie is great?)—No, it is ,just good/fine.

b) (You really like it, do you?)—Yeah, it’s 'just 'great/'fantastic/'perfect!

c) (How do you like it?)—It’s ,just good. vs It’s 'just 'great!

Additionally, the functions of just as used in certain specific collocations are explicit and stable. For example, in be just doing/going to do sth, be just about to do sth and other perfective-aspectual structures, just often has a specificatory meaning, indicating that something will happen soon or has just happened. It is used to lay emphasis in such collocations as just then, just in case, just any, right/ exactly just ²⁶ , and it usually means “minimally” when modifying some modal verbs, such as might, may and could, while it normally indicates “exactness” when modifying as, like, the same, etc. Needless to say, the meaning of just is much stabler and clearer when it is used in certain idiomatic expressions, such as It is just that…, I just feel/think/can’t…, I was just asking…, just a minute/moment, just see/hear something and just about (Cf. Aijmer, 2005a ). Therefore, the matchup between such forms and functions showss that certain usages of just demonstrate some constructional features. Molina and Romano (2012) have proven their psychological reality. Both Wierzbicka (1991: p. 351) and Aijmer (2005a) point out that just, though it goes through a change from expressing propositional content through expressing subjectification in interactions to functioning as a pragmatic marker, is very likely to develop further pragmatic functions; and, with high frequency of use and under social conventions, some of the functions can become increasingly concrete and stable. Thus, it is arguable that the constructional features of just show that it has acquired certain polysemous nature.