Compensation: A Synchronic Dynamic Mechanism in Language

Zhiguang  Abstract: In this paper, compensation mechanism—a synchronic dynamic rule in language is postulated and theorized. The rule raises the acceptability of clauses by effecting the positive side of some variations with exceptional expression and offsetting their negative side. This mechanism is determined by the instrumentality and systematicality of language so that its basic principle should be applicable to all exceptional expressions for particular positive effects and thus


Introduction
Compensation is one of the synchronic dynamic mechanisms, used in self compensation for inadequate clauses. It raises the acceptability of clauses by effecting the positive side of some variations with exceptional expression and offsetting their negative side. ① Compensation aimed at emphasis (compensation for short) is one of the most frequent and important acts of compensating. This paper is focused on this subject with an analysis from six aspects, in an attempt to prove a universal mechanism of synchronic compensation: 1. acceptability and statistics, 2. preliminary compensation, 3. advanced compensation, 4. the determinative role of instrumentality, 5. systematic decision, 6. universality.

Acceptability and statistics
Acceptability is the extent to which linguistic data would be judged by native-speakers  Ma Qinghua: School of Liberal Arts, Nanjing University, Nanjing, 210093, China. E-mail: maqh86@ yahoo.com.cn.  Wang Zhiguang: School of Foreign Languages, Zhejiang University, Hangzhou, 310058, China. E-mail: wazhiguang @yahoo.com.cn. ① Compensation is different from supplementation, or a strategy to make up for semantic losses, e.g. -dōngli tài lěngde hěn (冬里太冷得很, too cold in winter)‖ in which -tài (太, too)‖ is a supplement to the degree of emphasis in -hěn (很, very)‖ in Northwestern dialects (Long Guofu, 1958), and from substitution, or substitution of an apparently irrelevant form for an absent form in a system (Ma, 2005[2]:406), which comes from biology. Wang Jue (2001) used compensation in the sense of substitution. (cf. Wang Jue, 2007). to be possible in their language (D. Crystal,1985, p.2). This is measured by addressee's psychological acceptance, which varies from person to person because it involves not only static linguistic structure and its rules but pragmatic elements in dynamic application, for instance, relevant extralinguistic knowledge, social background (dialect), psychological background (attitude), physiological background (age, sex) and register. So the acceptability of clauses is a socio-psychological variant.
Acceptability and adequacy are different but mutually complementary as indispensable categories of the analytic compensatory mechanism. Adequacy refers to the conformity of clauses to the static systems (including phonology, grammar and vocabulary, etc.) and their rules. ① As far as compensation mechanism is concerned, recourse to acceptability to the neglect of adequacy means inevitably analysis of communicative strategy.
Clauses are divided into norms and variants according to adequacy, and are effective or ineffective according to expressive effect. Clauses are effective if the addressee understands and the speaker can expect their effects. ② There are four types of relationship: 1) [+normal +effective], 2) [+normal -effective], 3) [-normal (variant) +effective], 4) [-normal (variant) -effective]. Acceptable clauses exist in 1) and 3) and compensation occurs in 3). Clauses in 4) are not acceptable, while case 2) only occurs in abnormal communication.
In current practice, acceptability can only be differentiated on three levels: totally acceptable (unmarked), marginally acceptable (marked -?‖), or totally unacceptable (marked -*‖). This differentiation is not adequate in describing minor fluctuations on the same level of acceptability. To be more precise, it is feasible to decide acceptability by judging the hesitancy in accepting, namely 1 degree (simply judged acceptable), 0.75 degree (tending to be judged acceptable but with some hesitancy), 0.5 degree (always hesitant), 0.25 degree (tending to be judged unacceptable but with some hesitancy), 0 degree (simply judged unacceptable). See the following scale: Scale of Acceptability of Clauses 0.75 0.25 0.5 1 0 We used this scheme to do questionnaires and statistics of almost all the samples in this paper with 54 Ningjing University undergraduates. We averaged their scores on the acceptability of each clause as the mean acceptability of that clause (E). The fluctuation I of the acceptability of clause Y against clause X is I= (E y -E x ) / E x. The result of the investigation is attached in the Appendix, as the possibly scientific quantitative evidence of the qualitative description given here.

Preliminary compensation mode
The basic condition of a specific clause (the sum of elements like vocabulary, clause type and meaning) determines the means by which it can be adjusted (eg. -chīle(吃了, eaten)‖ cannot be replaced by -chī chī chī le( * 吃吃吃了, * eat eat eaten)‖ or -chī le le le ( * 吃了了了, * eaten-en-en)‖). The expressive aim in turn decides which means to choose.
There is the formula for compensation:

Base (B) · Regulation (R)  Aim (A) ①
Thus, we get the preliminary and advanced modes of compensation as the case may be. Variants play different roles in the two modes: as adjusting strategy in the preliminary compensation, and as the basis for clauses in the advanced mode. Emphatic compensation provides four means of variation (cf. §5.2.adaption to register): 1) violating the principle of economy, 2) eliminating lexicalization, 3) violating regular syntactic rules, 4) violating common sense. There are two forms of aim, emphasizing generation (A 1 ) or elevation (A 2 ), sometimes with some other effects.
Preliminary compensation involves the adjustment of normal clauses into variants, to achieve a new emphatic effect, which helps to make a variant acceptable in the formula: B(normal clause) • R (X)A 1 (emphasis generated) There are the following situations: 2.1. X=Variant 1 Variant 1 means the redundant form becomes the means of semantic emphasis, with its positive effect aimed to achieve acceptability. For instance, common sense shows that -shǒu (手, hand)‖ as instrument of the action in (1a) and case marker -yòng (用, with)‖ are both implicit. In (1b), these are represented as tautology. Since it expresses emphasis, the clause is acceptable (as acceptable as (36a)). Again, overemphatic strong degree adverbs tend to cause emotions, strong enough when used alone. When superposed, its subjective emphasis will be detached from rational emphasis to offset its redundant negative side; thus making the superposition acceptable ② (Compare (2a) and (2b)). ① The complete formula should be -Need·Base·Strategy  Aim‖. Need analysis is part of complex condition analysis. Need or not is regarded as 1 or 0, presumably 1. ② -tàiguòyu (太过于, too)‖､-tàiguò (太过, too )‖ are basically lexicalized. -tà i guòfè n (太过分, excessive)‖ is an adjectival phrase, without usage of superposition.

Advanced compensation
Clauses with advanced compensation are based on variants. The strategies of augmentation, word order, exophora (reference, contrast, objectivation, context, etc.), resonance (two elements intensifying each other with similar functions), and differentiation adjust the variants via changing information content, conditions and/or flexible understanding to achieve the expressive effect and raise its acceptability. The combinations of variants and adjusting strategies (see table below) form the left side of the equation. Most of the combinations are found, with a few being non-existent or questionable.

Change of information content
Adjusting strategies are used to intensify the variants and make the information content ① There are many samples for this but they are not emphatic.
full, thus differentiating the expressive function to highlight the affective effort, and raising the tolerance for redundancy ① and thus acceptability. 3.1.1. Increase of redundant items Increase of redundant items upgrades emphasis or transforms linguistic functions to express intense emotions, so that structures not emphatic or functional are rendered acceptable with enough emphasis and affective function, in the formula: B 1 (X) • R 1 (increase of redundant items)A 2 (more emphatic) + intense emotion X= coordinate clauses with the same subject In the principle of economy, coordinate clauses with the same subject show a strong tendency to omission, without which the clause may be ungrammatical or unnatural (i.e. (5a), (6a)). But when there are 3 coordinate items or more, with stronger emphasis, they will by themselves create a context of emotional repetition, thus raising their acceptability (i.e. (5b), (6b)). They smoke opium, they drank medicated wine, they patronized actresses, they flirted with women ‗They smoke opium, they drank medicated wine, they patronized actresses, they flirted with women.' X= synonymous coordinate clauses Synonyms are usually not juxtaposed for fear of redundancy, but to express extreme emphasis, the negative effect of the strange repetition would be offset. For example, the juxtapostion of 父亲爸爸 (father papa) is not common because the negative effect is not offset, i.e. (7a). But the juxtaposition of three or more items means intense emotion so that acceptability is raised, i.e. (7b). ). More synonyms juxtaposed means more redundancy, but the positive emphatic effect is boosted in a parabolic variation where it subsides drastically after reaching its peak. A few cases only allow two items (i.e. (34b)), and most frequently there are three items, sometimes even more (i.e. (9d)). So loose juxtaposed trinomials (with a pause) are more emphatic than corresponding tight binomials (without a pause), thus overpowering or offsetting the negative redundancy and raising acceptability (compare (8b) and (8a)).
This is your fault crime This is your fault, evil, crime ‗This is your crime!' ‗This is your evil!' Strong degree adverbs only represent rational attribution when used alone, and will not add emphasis logically when juxtaposed. Their second items in affirmative clauses only suggest a failed attempt to raise intensity and thus they are often regarded as negative repetition (in fact not intensive enough to spin off strong emphasis function), barely acceptable (i.e. (9a), (9b), (9c)). In 2001, we had a survey among three correspondence college Chinese-major classes in Wenzhou and Taizhou. The 210 informants were mostly school teachers, who tended to reject the repetition. But when more items were involved (i.e. (9d)), the clauses became emphatic enough to be disparate and show the emotion underneath, thus offsetting the negative effect of redundancy and raising acceptability. Some informants even judged them as exceptionally good sentences. ② ① Abusive, meaning ‗the speaker has fucked the addressee's mother'. cf. Sample 25. ② Negative juxtapositions like -bànde bù shífēn tài hǎo (办得不十分太好, not very so good) | gǎnqíng shang hǎoxiàng bù tài hěn shùnlì (感情上好像不太很顺利, not very so smooth romantically) | pí qì bù dà hěn hǎo (脾气不大很好, not very so good-tempered)‖. This merely boosts subjective emphasis, and combination with negation shows euphemism (Ma, 2006:102, 219).
The several characters in neon lights seemed very so appealing 'The characters seemed very appealing in neon lights' c. ? Qiá n fù zǒnglǐ jiǎng de yǐjīng fēicháng hěn qīnghu le.
Ever so much very so thank your friendliness, indeed 'Thank you ever so much for your friendliness, indeed!'

Resonance
The variants are the means of emphasis, boosting the affectiveness and raising acceptability by resonating with other strong means, in the formula:
He has a tender wife, virtuous wife 'He has a tender and virtuous wife.' b. Nǎ gè nánren bùxiǎng yǒu gè wēnróu de qīzi, xiá nhuì de qīzi. sometimes be used at the end of a sentence in resonance with other strong sememes to raise the acceptability of disparate forms. For example, in (13b) and (14b), emphatic disparate forms in resonance with strong sememes -měi tiáo (每条, each and every)‖ and -wánquán (完全, completely)‖ reinforce the emphasis and emotion to raise the acceptability of (13a) and (14a) (15a) and (16a) with extraposed adverbials or attributives are not very emphatic ((15b), or questionable (16b)), but in coordination with repetitive markers, they become more emphatic and good sentences ((15c) and (16c)

Change of conditions
Clauses with variants can, by strategy, temporarily change their speech objectives to express meaning or effects not directly connected to variants, shifting over-strident attention from variants, which then survive to exchange for compensation with their positive emphasis.

B 1 (coordinate clause with the same object) • R 1 (with connetive marker)  juxtaposable + A 1 (emphatic)
The addition of connective markers sometimes brings unjuxtaposable redundancy together, to be emphatic under this condition. Example (17a) is ungrammatical because of the repeated object, without which -gàishang (盖上, cover), xiānkāi (掀开, uncover)‖ would be a consecutive imperative. With the coordinative marker, it becomes grammatical even with a repeated object (17b).
Cover cloth, uncover cloth sometimes cover cloth, sometimes uncover cloth ‗Cover cloth, uncover cloth' ‗sometimes cover cloth, sometimes uncover cloth'

B 1 (synoymous coordinatives) • R 3 (rhythmic symmetry)  musical + A 1 (emphatic)
Symmetrical rhythm can connect redundant combinations otherwise not juxtaposable, to be emphatic conditionally. For example, most informants felt (18a) to be wordy, but the same combination is grammatical and emphatic in (18b), where there is rhythmic juxtaposition.

Flexible comprehension
Absurd sentences will be judged unacceptable if spoken by beginners of Chinese. But native speakers of Chinese are trusted for their competence and people will try to compromise in understanding them on artistic level instead of on normal level. A change of mind makes an absurd sentence original and humorous, with cartoon-like positive effect, and thus more acceptable. Absurdity here becomes the means to differentiate the speech levels. Acceptable absurd sentences are usually the result of coordination of implicit contrastive strategy and other effective strategies. There is an implicit triangle between its real meaning, literal meaning and common sense, which is the basis for novelty and emphatic effect. The compensatory formula is:

3.3.1.X=R 1 (addition or subtraction of meaning)
There is reasonable addition or subtraction of meaning in the artistic comprehension of absurd sentences, and emphasis and humor are achieved by means of the implicit contrast among real, literal meaning and common sense. When (19) (20) is understood as a, the semantic features of the components are contradictory, so it is not acceptable. If understood as b, we only get the feature -dùzi dà (肚子大, potbelly)‖, and discard the gender (metaphor), and then it is humorous to say so; emphasis raises its acceptability. When (21) is understood as a, it is against common sense and unacceptable. As b, we only get the meaning of -X>Y‖ in -X shà ng fùdà i Y (X 上附带 Y, X carries Y)‖, and with the implicit contrast to common sense and literal meaning, we get exaggerating and humorous effect with the reversal of size to underline a large nose on a small face, making the sentence more acceptable.

X=R 5 (differentiating the nature)
The nature ① of the content of absurd sentences is differentiated to dissolve the meaninglessness, so that the heterogeneous elements do not conflict with it, while the coordinated contrastive elements become humorously obvious by the false semantic conflict, to make the clause more acceptable. Differentiation includes that of narration or reporting (22b), that of truth or falsity (23b), direct narration or metaphor (24). Examples (22a) and (23a) are unacceptable absurd sentences.

Comprehensive method
Several strategies can be applied in one or more channels in compensation to achieve a corresponding positive effect. The formula for compensation in extra emphasis is:

Multiple strategies in a single channel
① The nature of clause content at the top and lower right angle of the semantic triangle, in relation to interlocutor, and functional nature in the speech acts. ② Zhang Haidi is a disabled Chinese writer.
In term of adjusted information content: X= end synonymous juxtaposition, Y=R 2 (changed word order) + R 4 (resonance), Z=intense emotion Juxtaposed synonyms at the end of a sentence are unacceptable if the focus is on the former, with the affective effect not strong enough to offset the repetitive negative side (25a). Change of word order to focus on the latter, to form an escalating tendency, will resonate with the emphatic redundancy in intense emotion, so that (25b) is passable (the subjectivity of the judgment sentence also participates in stimulating emotion).

X= strong degree adverb repeated, Y=R 1 (redundant items increased) + (R 3 (reference) + R 4
(resonance)), Z=intense emotion College and high school field inquiry yielded a dozen samples to prove: strong degree adverbs (except for the superlative and overemphatic adverbs) repeated once are barely acceptable (e.g. (26a) and (27a)), but repeated more times can be more acceptable via effective strategies, or even form extraordinary good sentences (e.g. (26b) and (27b), in which an astronomical figure brings explanatory reference and resonance). This is similar to juxtaposed strong degree adverbs (cf. above).

Combination of various strategies in multiple channels
Combination of two or three channels in condition, information content and flexible understanding will raise the acceptability of clauses through joint forces. There are the following situations: X= sentence-end compound verb disparate, Y=R 5 (differentiation) + R 3 (objectivation) + R 4 (resonance), Z=intense emotion Disparate use of non-disparate verbs can raise acceptability by the following means: 1) Conditioning. Disparate use by objectivation, in which a segment of language is temporarily transformed into a reported unit (similar to a quote) in another segment (Ma, 2005[2]:54). People are not so strict with quotes in judging grammaticality. The reported unit forms an exophora to explanatory elements. Variants in disparate use -shuàiguo lǐng (率过领, took the lead)､tǎoguo lùn (讨过论, discussed)､ là ngguo fè i (浪过费, wasted)‖ in (28b)-(30b) are all objectivized as quotes, so that they are more acceptable. In deep structure, emphatic clause type or emphatic adverbs are descriptions of them, so this type of reference is the semantic basis for objectivation. 2) Adjusted information content. It is noted that the positive effect of (28a)-(30a) disparate use is not enough to offset the negative variants, so that the clauses are not acceptable. So the explanatory elements to support the variants in (28b)-(30b) form a more extensive emphatic situation, which, in resonance with disparate use ① , forms a higher degree of emphasis to offset the negative side.  X= absurd sentences, Y=R 3 (contrast) + R 1 (changed understanding) + R 4 (multi-meaning in synergy), Z=novelty Absurd sentences can rise in acceptability by the following means. 1) Conditions changed. Some absurd sentences can contrast with familiar proverbs expressing extreme statement (32a), and become more emphatic and grammatical (32b). 2) Flexible understanding. Pushing the impossible in clause meaning (32b 2 ). 3) Changed information content. Just add up emphasis from two different sources to exaggerate and impress, forming exceptional acceptable good sentences (32b 2 ). Without contrast with familiar proverbs for emphasis, acceptability is low ((32c) and (32d)).

The decisive role of instrumentality of language
The principles and features in a language are all derived from limited preliminary features and the compensation mechanism is no exception. Its formation and operation are governed by instrumentality and systematicity. Language is an audial ① tool for exchanging knowledge, feeling and meaning created under the condition of mental and physiological acceptability. Instrumentality dictates that language must be effective and optimized, by which compensation is feasible, rational and implemented.

Principle of effectiveness
If a language is to be an effective tool to express and exchange, it must be as practical as all tools, and deeply utilitarian. When rules conflict with effectiveness, the result tends to favor the latter. Compensation then comes in.
In a moment of excitement, social restraints are resolved partially so that inspiration and creativity prevail, with innovative speech gushing out. ② It is clear to the speaker that special effects will come only out of breaking the ordinary rules, with formal or meaning ① Later extended to visual. ② Affective motivation is a major momentum for lexical change (H. Sperber, 1929), and a means of innovative use of language. breakthrough to realize the affective function. That is why the emphasis in compensation is mostly affective.
Variants are tolerated only when there is the expected effect, to realize their exceptional value. Variants that bring no positive effects will be rejected by the language system as falsity (i.e. (33b)). Compensation is obviously only utilitarian compromise against rules.
He meal already ate He meal already ate ‗He has had his meal.' ‗He has finished his meal.'

Principle of optimization
Any tool should be versatile, and language is being optimized all the time. This leads to a basic maxim, maxim of quantity, which governs the adjustment in compensation: speakers' contributions should be as informative as is required for the current purposes of exchange, and should not be unnecessarily informative (Crystal, 1985:190).
The maxim of quantity establishes the function between positive effects and acceptability. 1) Adjustment and compensation yield special effects to abundantly offset the negative side, so that variant clauses become exceptionally good sentences. 2) Adjustment and compensation yield special effects to offset the negative side, so that variant clauses become acceptable. Most of the sentences noted in this paper have achieved 1) and 2) in compensation. 3) Adjustment and compensation strengthen positive effects to counter the negative side, so that variant clauses become somewhat more acceptable ((16b) and (17b)). 4) When variants reach peak positive effects, attempts to further extend them will be contained by maxims of economy, and acceptability is reversed. There are the following three samples.
Sample 1. If two strong enough synonyms are juxtaposed (34b), when we add another strong synonym (e.g. (34a)), no emphasis will appear. On the contrary, tautology will make the clause less acceptable (34c). Sample 2. Strong degree adverbs when further repeated will be contained by the maxim of economy. For example, the superlative -zuì (最, most)‖ repeated once is enough to express emphasis (35b), but if repeated more times it means less acceptability. Sample 3. When common sense is repeated to exaggerate its emphasis (36a), the redundant negative factor is increased if the logical stress is on it and the sentence becomes less acceptable (36b).
Do not with hand pinch person (grammatical stress) Do not with hand pinch person (logical stress) ‗Do not pinch me.' ‗Do not pinch me with your hand.' So there is the function between positive effects and raised acceptability, which is the necessary result of language optimization.

System decision
Possibility, rationalization and implementation are mostly decided by the language system except for direct influence of instrumentality. Of course, the system is fundamentally determined by the instrumentality. If a language is to express the infinite open and changeable world of human knowledge, feeling and meaning, it has to be a complicated system of innovation, development and flexibility. So compensation is characterized by its self-stability, register adaptation and factor interaction, all representations of the language system. 5.1. Self-stability Any active system will sustain and stabilize itself. A language function lost can be -retrieved‖ with certain interfering variants under new conditions. Such fluctuations can be repeated time and again on different basis, for different objectives, by different strategies, so that a simple language can become more complex (Ma, 2005[2]:9,395-396). Compensation is just one of the retrieval processes in the system. Compared with retrieval without meaning compensation, its self-stabilizing effects can be better understood. The former is a normal fluctuation, a restraint and release of the reasonable potential of a language, or self-discipline before monopoly, intentional rejection of reasonable potentials which might disrupt the system. But, with interfering variants, the restraint is released and the usage becomes acceptable. Take juxtaposition of monosyllables. Though in theory coordinate phrases presuppose flexible word order, and can appear in any position ① , two coordinated monosyllables are actually not allowed in almost all positions ② , unless they are to be lexicalized or pre-ordered ③ . Disyllables dominate contemporary Chinese vocabulary, most of them are coordinate compounds (Zhang Dengqi, 1992). Once two monosyllables are put together, the addressee will assume lexicalization (any pause in between will be resolved by rhythmic pressure). This assumption is contradictory to the linguistic fact, so that people are repulsive to the latter, judging the whole clause as ungrammatical ((37a)-(41a)). This restrains the otherwise legitimate use. But under the adjustment of interference (with added markers in (37b) and (38b), with downgrading element in (39b)) and extraposition (objectivized in (40b), implicit contrast in (41b)), the clause will resume its adequacy, becoming more acceptable.
(37)a. xiǎng wǒ mā､ diē b. xiǎng wǒ mā hé diē Non-compensatory retrieval is also a case of self-stabilization. It has no extra positive effects, because the released potential is reasonable in any case. The fluctuation of acceptability is a witness to the self-stability of compensation.

Register adaptation
Systematic adaptability requires adequacy in a language, where a clause must be adaptable to style. Various functions for diverse registers call for different expressive styles to serve the register, which restrains the style to maintain its existence, so that style requirement also directly influences acceptability. For example, scientific register is realistic, while poetry is given to bizarre imagination. Absurd sentences are not found in scientific discourse because they deviate drastically from logical truth, but abound in poetry, where they arouse dead images in language as very poetic, visionary and imaginative means to express originality. Chomsky's nonsense sentence -? Green ideas sleep furiously‖ is said to have been found in poetry (with certain devices, poems with such a sentence are not impossible), not as Hu Mingyang (2000) said, -Any English native speaker would not think such a sentence acceptable‖. ① ① Liu Dawei (2001: 206, 216) studied absurd sentences in literary language. Liu Dawei & Wang Zhiguang (2007) put forward the possibility of features, in a further probe into this.
No register rejects emphasis. If spoken unscrupulously, any clause can be emphasized by certain kinds of stimuli. It is not hard to understand that to emphasize, we must try to impress addressees, so we should get formal or semantic stimuli. The following formal augmentation by simulation can cause empahsis: {stress, elongation (e.g. long vowels), disparate use of set phrase (-Chaici (拆词) ① ‖ etc.), juxtaposed synonyms, repetition, extraposition (added position for elements)}. Semantic (in addition to formal) methods of emphasis include antithesis, rhetorical question, redundancy, absurd clause, selection of eye-catching synonyms ② . Some methods are normal (stress, elongation, rhetorical question, antithesis, selection of eye-catching synonyms); others are variants (disparate use of set phrase, extraposition, semantic absurdity, redundancy (superposition, repetition ③ )). Variants alone involve compensation. Emphatic compensation exchanges variants for positive effects, as a means to form vivid artistic register or everyday casual style, so it is not found in typical scientific texts.

Interactive elements
The language system dictates the purpose of communication to organize items of linguistic structure, to adjust their interrelationship, and to suit structure to its environment, so that within the structure only this movement instead of that is allowed among various possibilities. So, in all the compensation modes, there are interactive relationships among the foundation, adjustment, object and raised clause acceptability. There are three types of interaction. The first type is mutual, based on the same need, which contains the effect of A on B and B on A. The factors in a resonance all serve the emphatic expression, so that clause intensity is obviously raised. Compensation only occurs in vivid registers or casual speech, which need it to enliven the atmosphere. The second is based on different needs, so that the effect of A on B is different from that of B on A. For example, variants often carry exceptions expressing exaggeration or special interest, which in turn raises clause acceptability. The third type is directional, that is, without the effect of B on A. Variants of the advanced compensation and adjustment strategy in condition change exemplify this type, with the latter providing the former with favorable environment in which to express positive emphasis (cf. §3.2. Change of conditions).

Universals in language
① -Chaici (拆词)‖ is a figure of speech, the disparate use of set phrases is aimed at a meaning. ② -yì chá ng (异常, unusual)‖ is more emphatic than -fēicháng (非常, uncommon)‖ (Zhang and Zhang, 2005: 257). ③ Repetition is counted as a variant because it violates the principle of economy.
In this paper, the rules of compensation are investigated in the field of emphatic compensation. Analysis is made into the variants, which achieve positive effects via exceptions, in order to offset the negative effects of variants and raise clause acceptability. After discussing their root causes, we conclude: Compensation is the result of the instrumentality and systematicity, rather than an accidental phenomenon; language must compromise against existing rules under utilitarian conditions, and follow informational adequacy in compensation; there are interactive relationships among clause bases, adjustment, objective and the raised acceptability. Fluctuation in acceptability is self-stabilizing, and follows the principle of adequacy in compensation. This is the concrete principle of operation in compensation. Since compensation is directly determined by the instrumentality and systematicity, its basic principles not only cater to the expressive activity of emphasis, but also to linguistic exceptions for other purposes (e.g. poetic license as variant), thus compensation is a synchronic dynamic mechanism that is universal. This is the final conclusion of this paper, which shows the value of our academic endeavor.