2.4　Pragmatic transfer，pragmatic failure and fossilization

In SLA literature，there has been ample evidence of the complicated and controversial role of L1 in L2 learning process，no matter whether it is termed as interference and transfer，which are associated with behaviourist theories of L2 learning，or a theory-neutral term crosslinguistic influence(Kellerman & Sharwood-Smith，1986).So is the case in the field of ILP.The influence of learners' L1 pragmatic knowledge on their L2 pragmatic performance has been a focal concern in ILP studies(Kasper，1981，1992；Takahashi & Beebe，1987；Takahashi，1996；Maeshiba et al.，1996，for instance)in spite of the widespread recognition that transfer is more conspicuous at the level of phonology，lexis and discourse(Ellis，1999a/b/c).In order to examine the effect of Chinese ethnic learners' ILP development，a selective review of what have been done so far will be helpful.Pragmatic transfer in literature is first of all introduced；a brief review of fossilization studies follows，and that of pragmatic failure comes in the end.

2.4.1　Pragmatic transfer

Etymologically，the term transfer in Latin means“to carry”，“to bear”or“to print”，“to impress”or otherwise“to copy”(as a drawing or engraved design)from one surface to another (Webster's Third New World International Dictionary，1986：2427).It is a pervasive phenomenon，not only in language learning but in all learnings.As Ausubel put it in the motto of his book Educational Psychology：A Cognitive View：

If I had to reduce all of educational psychology to just one principle，I would say this：The most important single factor influencing learning is what the learner already knows.Ascertain this and teach him accordingly(1978：iv).

Foreign language learning situations involve NL transfer，or more generally，cross-linguistic influence.Previous knowledge can facilitate learning(positive transfer or facilitation)to the extent that the established knowledge and the“new”are identical and no effort therefore need be made to learn the“new”(which is not new，since it is already known).But established knowledge can also interfere with learning(negative transfer or interference)，either by leading to inappropriate behaviour in the new context (intrusive interference)or by preventing or inhibiting the learning of the new(inhibitive interference)(cf.Hammerly，1991：64).Furthermore，crosslinguistic influence affects typologically distant or close target languages and their components in different ways.

Most studies on languagetransfer in the last three decades have focused primarily on the influence of the native language on a second language or interlanguage.Despite many years of intense discussion on the phenomenon of language transfer，researchers have rarely focused their attention on instances of transfer from languages other than the native language.This fact is clearly reflected in the various definitions of language transfer or“crosslinguistic influence”(Sharwood-Smith & Kellerman，1986)in literature.The term crosslinguistic influence is generally used as a superordinate term，thus including instances of native language transfer，interlanguage transfer，avoidance due to influence of another system，and even“reverse transfer”from an interlanguage back into a native language.In terms of“transfer”itself，Odlin，in the most general book on the subject，described it as“the influence resulting from similarities and differences between the target language and any other language that has been previously(and perhaps imperfectly)acquired”(2001：27).Gass and Selinker，in their introductory textbook，provided a glossary of key terms based on the most common use，concluding that“for most researchers，language transfer is the use of native language(or other language)knowledge—in some as yet unclear way—in the acquisition of a second(or additional)language”(1983：372).Along the same line，Sharwood-Smith described his view of“crosslinguistc influ-ence”as“the influence as the mother tongue on the learner's performance in and/or development of a given target language；by extension，it also means the influence of any‘other tongue’known tothe learner on that target language”(1994：198，italics in the original).

What these definitions have in common is the inclusion of anonnative language as a potential source of transfer.In other words，they imply the transferability of both native and nonnative knowledge in interlanguage production.But it turns out that there is at present no evidence in the language transfer literature for such a claim except for Angelis & Selinker(2001)，who explored the detail of interlanguage transfer[2]，setting the stage for more comprehensive and detailed empirical work.In brief，based on the assumption that all linguistic systems present in the speaker's mind may be simultaneously interacting and competing in interlanguage production，their research revealed that“current language transfer theories are highly restricted”and“cannot be comprehensive if its principles are based on two languages only”(2001：44).Coincidentally，such a view is also our concern in the present research.

When it comes to pragmatic transfer，Kasper defined it as“the influence exerted by learners' pragmatic knowledge of languages and cultures other than L2 on their comprehension，production，and acquisition of L2 pragmatic information”(1992：207).The condition of transfer occurrence is technically referred to as trans-ferability(Kasper，1992).Unlike the research in the days of the Contrastive Analysis Hypothesis，SLA research on transferability currently takes a cognitive approach to transfer.Kellerman (1981)，for instance，viewed the learner as deciding for him/ herself which forms and functions of the L1 are suitable for use in the target language.And he found three factors which affect trans-ferability，namely：

(1)psycholinguistic markedness，i.e.whether a learner regards a particular feature as unique to his/her L1；

(2)the reasonable entity principle(REP)which involves learners' belief as to what is possible in the L2 system；

(3)psychotypology which represents learners' beliefs regarding the distance between the target language and another language，usually the L1.

——Barron，2003：39

Based on these factors，ample interlanguage pragmatic researches have been conducted to test their effect on pragmatic transfer (cf.Fouser，1997；Cohen，1997；Olshtain，1983；Robinson，1992；Takahashi，1993，1996).In addition，Takahashi observed three dimensions of transferability in SLA research，namely，the study on the developmental sequence of transferability，lin-guistic markedness，and non-surface form transfer(1995：11).And these three dimensions，Takahashi further indicated，are usually approached in ILP research from the perspective of psycholinguistics(Kasper，1981；Olshtain，1983；Olshtain & Cohen，1983，1989；House & Kasper，1987；Bodman & Eisenstein，1988；Robinson，1992；cited in Takahashi，1995：46 -49).

In her“Transferability of indirect request strategies”，Takahashi(1992：69 -124；1993：50 -83)made an explicit investigation on transferability of five indirectness strategies realized by the“conventions of usage”of Japanese indirect requests when Japanese learners of English realized English indirect requests in four situations.Takahashi used informants representing two proficiency groups：beginning-intermediate level students，considered as low ESL group，and advanced students，high ESL group.They were presented with an acceptability judgment task for five indirect request expressions in respectively Japanese and English for each situation.The transferability rate was computed for each strategy for each situation by subtracting the acceptability rate of the English indirect request from the acceptability rate of the corresponding Japanese indirect request.The obtained transferability rate was considered as the“psycholinguistic markedness”of each strategy，which determined its language-specificity or neutrality.This study manifested that contextual factors play a major role in determining transferability at the pragmatic level.The results also displayed that some proficiency affected the transferability of those indirectness strategies.Following the initial findings，further attempts were made to explore the type of contextual factors that were most likely to affect transferability，and to expound the relationship between the proficiency effects on the transferability of the indirectness strategies.

The transferability of the five indirectness strategies realized by the Japanese learners of English was further discussed in Takahashi (1995).The study showed that the Japanese learners differentially transferred the Japanese indirectness strategies.Furthermore，Takahashi detected that the transferability of each L1 request strategy was determined by the interaction between politeness and conventionality perceived in each strategy and the degree of mitigation required in each imposition context.In addition，the transferability rate was influenced by the proficiency factor.However，there was not a definite tendency for a positive correlation or for a negative one between L1 transfer and proficiency.

Depending on the source of transfer，pragmatic transfer can be in general divided into two types：positive and negative.The former occurs“when language-specific conventions of usage and use are demonstrably non-universal yet shared between L1 and L2”(Kasper，1992：212)；the latter occurs when“L1-based sociopragmatic and pragmalinguistic knowledges［are］projected onto L2 contexts and［differ］from the pragmatic perceptions and behaviors of the target community”(ibid.，213).Although the manifestations of language transfer can be errors，facilitation，avoidance and over-use(Ellis，1999a：302)，the traditional focus was placed on errors that learners produce，that is，the negative effect alone.Similar to transfer studies at other language levels，ILP studies have been much more concerned about negative pragmatic transfer，with positive transfer touched on slightly(Maeshiba，et al.1996；LoCastro，2003，for instance).The lack of research on positive pragmatic transfer may be due to methodological difficulties in the identification of positive evidence—to distinguish it from either universal pragmatic knowledge or general developmental features on the one hand，or due to the trend to study language phenomena which usually result in communication problems rather than successful effect which tends to go unnoticed and neglected on the other(Barron，2003：37；LoCastro，2003：255).

Still in terms of the ineffective communicative outcomes，the pragmatic transfer effects have been noted at the pragmalinguistic and sociopragmatic level according to the distinction made between pragmalinguistics and sociopragmatics by Leech(1983：10 -11)，which Thomas(1983)picked up in classifying the types of pragmatic failure.Pragmalinguistics，in Leech's definition，refers to our linguistic knowledge of language use，and sociopragmatics is related to how our sociological knowledge influences our interaction.And that is why Kasper remarked that“Leech's(1983)distinction between pragmalinguistics and sociopragmatics，applied by Thomas(1983)to identify two major types of‘pragmatic failure’，is equally suitable to broadly separate the two main loci of pragmatic transfer”(Kasper，1992：208).

Put it another way，pragmalinguistic transfer refers to the process by which learners select certain strategies and forms from their L1 in their interlanguage.In Kasper's words，it is“the process whereby the illocutionary force or politeness value assigned to a particular linguistic material in NL influences learners' perception and production of form-function mappings in TL”(Kasper，1992：209). Sociopragmatic transfer concerns learners' reference to their L1 perceptions of social relationships and contextual variables when deciding whether to perform a particular illocution or not.It is a process“operative when the social perceptions underlying language users' interpretation and performance of linguistic action in TL are influenced by their assessment of subjectively equivalent NL contexts”(ibid.).Despite the clear distinction by definition，these two concepts are interrelated and often difficult to identify in practice(Kasper，1992：210).The possible influence of pragmalinguistic and sociopragmatic transfer will be of course identified in the present study，but our discussion will include the different realizations of transfer effect such as facilitation，over-use or avoidance which might lead to other communication consequences than communication failures.

Accordingly，it can be inferred from Kasper's dichotomy of pragmatic transfer that negative pragmatic transfer also has two corresponding types.The first type is negative pragmalinguistic transfer，and the other，negative sociopragmatic transfer.

Kasper's dichotomous treatment of pragmatic transfer seems to have embodied two considerations.First，by putting some of the pragmatic transfers under the category of pragmalinguistics，due consideration has been given to the internal linguistic features of a pragmatic transfer，and similarly，an inclusion of sociopragmatic transfer has given due consideration to the social or communicative feature a pragmatic transfer takes.That seems to be a sound treatment and that is why her classification of pragmatic transfer，positive and negative，has never been challenged since 1992 and in fact her classification has been taken as a framework in pragmatic transfer studies.

Negative pragmalinguistic transfer has been documented in a host of studies focusing on mainly the three speech acts of request (Blum-Kulka，1982，1983；House & Kasper，1987；Faerch & Kasper，1989；Takahashi & Dufon，1989)，complaint(De-Capua，1989)，and apology(Cohen & Olshtain，1981；Olshtain，1983；Trosborg，1987；House，1988；Beebe et al.，1990；Bergman & Kasper，1993).

Mother tongue-based pragmalinguistic forms such as the“can you”question form were observed to be transferred into the making of an ability question(“can you”)，displaying the learners' inability to convey the pragmatic force.Besides，mother tongue directness strategies were also transferred into the making of a request，manifesting the learners' NL pragmatic preference in using indirect strategies in situations where the Hebrew as a TL demands more directness(Blum-Kulka，1982，1983).A similar observation was made among the Danish and German learners of British English(House & Kasper，1987)and the direct German linguistic forms in realizing the speech act of complaint in English(De-Capua，1989).Further reports showed that the level of directness or indirectness is associated with the level of desire for something among the Japanese(Takahashi & Dufon，1989；Beebe et al.，1990).In cases with a strong desire for something，the Japanese depend on more direct strategies than the Americans do，while in cases when a desire is implicit，they incline to less direct request strategies than the Americans(Takahashi & Dufon，1989).

Then over-use of the“excuse me”expression was found among the German learners，showing a high rate of German influence (House，1988).In a cross-cultural study(Olshtain，1983)，it was found that the frequency differences of apology making are presented with the English the highest，followed by the Russians and the Hebrews the lowest.However，the Japanese influence in the frequency of“excuse me”among the Japanese learners was found to be subject to situations and certain variables.In terms of time，place and parties，the Japanese learners use less pleadings than the Americans do，while in making refusals，the Japanese use more pleadings(Beebe et al.，1990).All these reports displayed that nonnative speakers' negative pragmalinguistic transfers took a semantic form mapping or a formal mapping.

Negative sociopragmatic transfers were also found to operate in Venezuelan politeness styles into Spanish(Garcia，1989)and Japanese status difference in making refusals(Beebe et al.，1990)were documented.The Japanese were found rejecting positive remarks and using formulaic expressions in politeness orientation(Takahashi & Beebe，1993)and modifying corrections by means of positive remarks and softeners(Takahashi & Beebe，1993).

In addition，various forms of NL-TL communicative style-mappings were detected.Scarcella(1983)showed that learners transferred from Spanish(NL)and German(NL)their acquainted communicative styles into English.These communicative styles were neither pragmalinguistic nor sociopragmatic transfers.Hence it seems that Kasper's division of negative pragmatic transfer needs expanding in order to incorporate those NL-based communicative styles.

Takahashi and Beebe(1987)proposed the positive correlationhypothesis，predicting that second language proficiency is positively correlated with pragmatic transfer.According to the hypothesis，low proficiency learners are less likely to display pragmatic transfer in their L2 production than higher-proficiency learners because they do not have the necessary linguistic resources to do so.Hill's study (1997)would also appear to support the positive correlation hypothesis.His highest proficiency learners appeared to be moving toward native-like request production；however，transfer appeared to be the result of first language influence.For example，advanced learners' use of conditionals such as if you do not mind is likely a direct translation from Japanese.This is also true of their overuse of apology moves(e.g. So I'm sorry very much，I'm feeling bad but and Sorry to interrupt you but)as external modifiers.It appears that Hill's more advanced learners transferred these forms to English from Japanese as soon as they had the requisite linguistic skills in the target language.

In short，if learners perceive great similarity between L1 and L2，they tend to transfer pragmatic norms and strategies from their L1.Robinson(1992)also suggested that learners may be more prone to transfer their first language pragmatic knowledge when they hold a universalist view as opposed to a relativist perspective on pragmatic norms.Cultural awareness is thus a factor in facilitating positive pragmatic transfer in ILP development.

2.4.2　Pragmatic fossilization

Similar to all the major concepts discussed above，fossilization is also one of the most important concepts in the field of SLA.Over the past 30 years，many researchers have conducted various studies on the phenomenon of fossilization.In most of their studies fossilization has been manifested phonologically，lexically and syntactically in the interlanguage system while research on pragmatic fossilization is left as a blank waiting to be filled.The main reason for the lack of attention is that more attention is given to the SL or FL learners' development of linguistic competence due to the great influence of structuralism.Thus，in this section，we will revisit the notion and features of pragmatic fossilization in ILP development so as to serve the purpose of the present research.2.4.2.1 The notion of pragmatic fossilization

The notion of“fossilization”can be traced back to scholars such as Weinreich(1953)and Nemser(1971)，who talked about “permanent grammatical influence”and“permanent intermediate systems and subsystems”(Han，2004：14)respectively.The term fossilization was first introduced into the second language acquisition literature by Larry Selinker in his paper“Interlanguage”published in 1972.According to Selinker，fossilization is both a cognitive mechanism known as the“fossilization mechanism”(1972：221)and a performance-related structural phenomenon.The two functions are thought to be interrelated as follows：

Fossilization，a mechanism ...underlies surface linguistic material which speakers will tend to keep in their IL productive performance，no matter what the age of the learner or the amount of instruction he receives in the TL(ibid.，229).

Fossilizable linguistic phenomena are linguistic items，rules，and subsystems which speakers of a particular L1 tend to keep in their IL relative to a particular TL，no matter what the age of the learner or amount of explanation and instruction he receives in the TL ...(ibid.，215).

Since 1972，the notion of fossilization has been expanded to“a permanent cessation of IL learning before the learner has attained TL norms at all levels of the linguistic structure and in all discourse domains in spite of the learner's positive ability，opportunity，and motivation to learn and acculturated into target society”(Selinker & Lamendella，1978：187).In accordance with the view expressed above，the scope of“fossilizable structures”is broadened from “linguistic items，rules and subsystems”to“all levels of the linguistic structure and in all discourse domains”.And fossilization in the sense of a general cessation of learning will culminate in ultimate“fossilized competence”：

Fossilization is the process whereby the learner creates a cessation of interlanguage learning，and thus stopping the interlanguage from developing，as is hypothesized，in a permanent way ...The argument is that no adult can hope to ever speak a second language in such a way that s/he is indistinguishable from native speakers of that language(Selinker，1996).

In brief，since 1972，Selinker has extended the scope of fossilization.On the one hand，the referential scope of fossilization expands from“backsliding”to“cessation of learning”and to“ultimate attainment”.On the other hand，the linguistic scope of fossilization also expands from fossilizable structure to a fossilized interlanguage.Therefore，the study of fossilization should not remain mainly in the fields of phonology，syntactics and lexicon；the study of pragmatic fossilization should also be included.

Over the past few decades，many other researchers have also given numerous interpretations of fossilization.For instance，Vigil and Oller suggested that fossilization applies to both incorrect and correct forms，but they still focused on grammatical fossilization，keeping pragmatic fossilization out of the picture.“We will extend the notion of fossilization to any case where grammatical rules，construed in the broadest sense，become relatively permanently incorporated into a psychologically real grammar”(1976：282).Rod Ellis(1988)also suggested fossilized target-like forms as well as fossilized errors.To some of the researchers，fossilization is a process as well as a product—“a process whereby repeated practice and exposure to the language does not lead to any further development”(Smith，1994：37).Han(1998：50)proposed a definition of fossilization as she took into account both the innateness and the external manifestation of the phenomenon below：

COGNITIVE LEVEL：Fossilization involves those cognitive processes，or underlying mechanisms that produce permanently stabilized IL forms.

EMPIRICAL LEVEL：Fossilization involves those stabilized interlanguage forms that remain in learner speech or writing over time，no matter what the input or what the learner does.

Like many other applied linguists，Han restricted the study of fossilization to the syntactic level in her early works.From the above definition of fossilization，we are still not sure whether or not she takes pragmatic fossilization into consideration.But in her recent works，where she discussed the future directions of fossilization research，she acknowledged that researchers should attempt to examine fossilization on a macroscopic and a microscopic level and that pragmatic fossilization is one aspect of fossilization in the learner's interlanguage.On the macroscopic level，researchers typically look at general failure across adult L2 learners.On the microscopic level，they look at individual learners and focus on the local cessation of learning that takes place in various interlanguage sub-systems such as phonology，morpho-syntax，lexicon and pragmatics(Han，2003：115).

As fossilization has become a widely accepted term，the definition of fossilization extends.The boundary of fossilizable items expands to phonological features，lexical and syntactic rules，pragmatic principles and communicative competence.However，despite the fact that more and more researchers tend to explain fossilization in a broad sense，studies on pragmatic fossilization are seldom seen in SLA literature，which is one of the chief deficiencies in current research on fossilization.

With the development of interlanguage pragmatics，some researchers begin to believe that the phenomenon of fossilization also applies to the pragmatic aspects of language development.For example，Thomas(1983)suggested that pragmatic competence cannot be simply“grafted”onto grammatical competence and that there may even be cases of pragmatic fossilization.

Trillo defined pragmatic fossilization as“the phenomenon by which a nonnative speaker systematically uses certain forms inappropriately at the pragmatic level of communication”(2002：770).And in his view，learners of a foreign language follow what is called a“binary track”in their linguistic development：the formal vs.the pragmatic track.The formal track relates to the grammatical and semantic rules that conform to the competent use of a given language；the pragmatic track，on the other hand，relates to the social use of language in different contexts and registers.Native speakers of a language would develop both tracks simultaneously by means of contact with natural language，and thus would establish a mutual relationship between both communication tracks.ESL/EFL learners in a non-target language environment，however，would develop the formal and the pragmatic tracks through formal instruction.The difficulty，therefore，is that the pragmatic track，linked to the cognitive，affective，and socio-cultural meanings expressed by language forms，is difficult to implement in educational syllabuses.In fact，the development of pragmatic competence demands a(pseudo-)natural foreign language context that is often almost impossible to produce in formal education.As a result，ESL/EFL learners find the formal-functional dichotomy problematic in their learning process due to the decontextualized nature of teaching environments.For instance，Bardovi-Harlig and Dornyei (1998)showed that their ESL group identified more pragmatic errors and evaluated them as more serious than grammatical errors.Besides，during the learning process，ESL/EFL learners are usually exposed to what has been called“semantic distance”between two languages—the principles that organize semiotic meanings in a language(Hasan，1984：193)，and“cultural distance”—the cultural and conceptual structure that models meaning in a language (Kecskes & Papp，2000：96).However，ESL/EFL learners are not exposed to what Trillo calls“pragmatic distance”(Trillo，2002：771)，the variants in the social，cognitive and contextual dimensions of linguistic communication that govern and systematize social relations in speech.In fact，pragmatic distance is an essen-tial factor in communication，as it models the relationship that exists between two speakers of a language(Kasper & Blum-Kulka，1993).

Due to these factors，persistent pragmatic errorstake place in the process of input as well as in the process of output.Therefore，Trillo addressed the persistent pragmatic errors as“pragmatic fossilization”and pointed out that pragmatic fossilization appears in ESL/EFL learners' systems“not because of a lack of competence in other linguistic areas such as lexis，grammar，etc.，but because there is a delay in the presentation of the pragmatic variation that exists with respect to the way communication competence is acquired in the mother tongue”(Trillo，2002：771).The phenomenon of pragmatic fossilization is one of the main problems that nonnative speakers of English face in their learning process.Thus Trillo suggested that the role of pragmatic fossilization should be regarded as an element to consider in foreign language learning and teaching.

There are also some studies that report on the empirical evidence for the existence of pragmatic fossilization.For example，Scarcella (1983)examined the variant use of conversational features by highly proficient Spanish learners of English who have arrived in the U.S.before age seven and have resided in the community for 12 - 17 years.As these deviations appear frequently，systematically，and consistently among the learners，Scarcella claimed that this inappropriate use of conversational features results from these learners'“wholesale transfer”(ibid.，319)at an earlier stage of development and has remained fossilized ever since.Likewise，Rehbein(1987)examined the use of discourse markers by Turkish immigrants learning German in a German industrial workplace and found that the overgeneralized，multiple-purpose use of one particular fixed routine does not decrease with extended years of residency in the L2 community.Although the formula in question is highly frequent and salient in the available NS input，more experience and input in the same community would not have helped these learners to defossilize.These two examples suggest that learners fail to attain an adequate level of sociocultural competence in spite of prolonged and frequent exposures to the L2 and that an ESL setting itself does not provide an adequate learning context for destabilizing fossilized interlanguage performance.

Fossilization has also been observed in the use of routine formu-las，appearing in the oversimplification in structure and overgener-alization in function of the routines in question.For example，Trosberg(1995)reported a fossilization phenomenon in the use of formulaic devices for polite requests(internal，syntactic modification and external modification by lexical down-graders)by highly advanced Danish EFL learners with more than ten years of instruction.Similarly，Takahashi(1996)claimed that after seven to ten years of formal instruction in English，her Japanese learners manifest a fossilized false perception of the functional equivalence between L1 and L2 request strategies.She further argued that such a false projection of L1 form-function mapping onto corresponding L2 linguistic categories results from transfer of training which leads the learners to assuming that the English modals would and could convey the same communicative effect for politeness as the Japanese honorific auxiliary verbs.In both Takahashi's and Trosberg's studies the learning context was foreign language learning.To sum up，all these findings，both in the ESL and in the EFL settings，seem to prove the existence of fossilization at the pragmatic level.

In China，researchers like Huang & Zhang(2004)and Li (2004)investigated the phenomenon of fossilization at the pragmatic level among advanced English learners and suspected that pragmatic fossilization exists among Chinese English majors.Zhang (2005)conducted two surveys in the foreign language teaching environment and drew the conclusion that the IL forms concerning language use in the FL environment often present a manifestation similar to fossilization.He proposes the term“pseudo-fossilization”to name such a phenomenon(ibid.，88).Huang & Zhang，after making an investigation among English majors，concluded that students come to a state of pragmatic fossilization in their second year of college education.Therefore，they held the belief that pragmatic fossilization is one of the demonstrations of IL fossilization(2004：88).Li(2004)compared the application of English discourse markers between Chinese EFL learners and native speakers of English and suggested that Chinese EFL learners are in a state of pragmatic fossilization in learning to use discourse markers and that pragmatic fossilization poses a problem to be tackled in foreign language learning and teaching.2.4.2.2 Features of fossilization

Many researchers have claimed that fossilization presents several properties，among which persistency，resistance and permanency are the dominant ones(Han，2003).

By persistency is meant the persistent failure to achieve complete mastery of the TL.For instance，Mukattash(1986)described fossilization as“persistent non-target-like performance”，and Schouten(1996)named it“structural persistence”，while Hawkins(2000)defined it as“persistent difficulty”.It is proposed that conscious efforts to change the plastic IL development are often fruitless because of this feature of fossilization.

By resistance is meant that fossilizable structures are resistant to external influences.In The Unabridged Random House Dictionary，“fossilize”is defined in the following way：

Ling.(of a linguistic form，feature，rule，etc.)to become permanently established in the interlanguage of a second-language learner in a form that is deviant from the target-language norm and that continues to appear in performance regardless of further exposure to the target language(Flexner，1993：755).

Permanency refers to the everlasting existence of erroneous forms.Brown defined fossilization as“the relatively permanent incorporation of incorrect linguistic forms into a learner's L2 competence”(1996：217).Also，in the Routledge Dictionary of Lan-guage and Linguistics(Bussman，2000：171)，fossilization is defined as the“permanent retention of linguistic habits，which，when taken together，constitute a learner's interlanguage.”

However，with the development of SLA，many researchers have argued that“fossilization is internally determined，due to the func-tioning of bio-cognitive constraints，but it can be modulated(i.e.aggravated or alleviated)by environmental，social，and psychological forces”(Han，2004：10).In other words，persistency，resistance and permanency are properties of fossilization at the cognitive level.It means that L2 learners lack the ability to attain native-like competence no matter whether they are child L2 learners or adult L2 learners.However，at the empirical level，i.e.in the respect of external manifestations，fossilizations can be reduced or avoided.Although maturational constraints and native language influence are the major determinants of the general lack of success in L2 learning，the degree of such a lack of success varies from individual to individual，due to the functioning of such variables as instruction.For instance，Ellis claimed that formal instruction helps to prevent fossilization：

Learners will fail to acquire the more difficult rules(e.g.inversion and verb-end)once they have achieved communicative adequacy.Learners may need form-focused instruction to make them aware of grammatical features that have little communicative importance and yet constitute target language norms.In other words，formal instruction serves to prevent fossilization ...naturalistic acquisition is often a very slow process；instruction may not alter the way in which learning takes place，but it may help to speed it up.(1988：314)

Ellis' claim can be and has been supported by empirical evidence.For instance，instruction can aid acquisition(Krashen & Seliger，1975；Long，1983)，and instruction is facilitative (Spada，1997；Norris & Omega，2000).Some studies prove that if there is no grammar instruction(including error correction)，fossilization will result(Higgs & Clifford，1982).In fact，few have questioned that instruction is a striking characteristic of adult SLA(Bley-Vroman，1989).Long(1983)，after reviewing 13 early studies of instructional effects，concluded that there is considerable evidence that instruction is beneficial(1)for children as well as adults；(2)for beginning，intermediate，and advanced students；(3)on integrative as well as discrete-point tests；and (4)in acquisition-rich as well as acquisition-poor environments (ibid.，359).

Norris and Omega(2000)，in a large-scale synthesis study of the effectiveness of L2 types of instruction，not only confirmed Long's findings but also made significant progress in terms of identifying different types of instruction.Their findings are summarized as follows：

(1)Focused L2 instruction results in large target-oriented gains.

(2)Explicit types of instruction are more effective than implicit types.

(3)Focus on Form and Focus on Forms interventions result in equivalent and large effects.

(4)The effectiveness of L2 instruction is durable.(ibid.，417)

In short，properties of fossilization at its cognitive level demonstrate unattainability of native-like competence for ESL/EFL learners.However，in the respect of external manifestations，fossilization can be alleviated，especially through instruction so long as instruction can aid second/foreign language acquisition.

2.4.3　Pragmatic failure

Researches from the acquisitional perspective have found that internal factors such as interlingual influence，pragmatic overgeneralization and teaching-induced errors，and external factors such as learner-specific input，learning context，and the like，might result in pragmatic failure，“the inability to use language effectively and to understand what is meant by what is said”(Thomas，1983：94).Thomas noted that pragmatic failure has occurred on any occasion“on which H(the hearer)perceives the forces of S(the speaker's)utterances as other than S intended he or she perceives it”(ibid.).

There is considerable room for intercultural misunderstanding (Levinson，1983：376).Both grammatical error and pragmatic failure may inhibit or hinder communication.As Thomas(1983：97)put it，“while grammatical error may reveal a speaker to be a less than proficient language user，pragmatic failure reflects badly on him/her as a person”[3].He(1988)also concluded that in interacting with foreigners native speakers tend to be tolerant of errors in pronunciation or syntax but violations of rules of speaking are often interpreted as bad manners.It is true that to a certain degree，pragmatic failure is much more serious than grammatical errors.This is because pragmatics—the study of language in use—is the field where a speaker's knowledge of grammar comes into contact with his/her knowledge of the world.It relates to whether an interlocutor speaks in a suitable manner，at an appropriate time，in accordance with the native speaker's way of speaking，etc.，as illustrated in Figure 2—1.It is thus worthwhile to make an effort to explore pragmatic failure that is actually an indication of low pragmatic competence or intercultural communicative competence.

Centering on the purpose in the present study，we will review three types of pragmatic failure，i.e. pragmalinguistic，socioprag-matic and pragmatic behavioral，based on Thomas' contribution (1983)and the inspiration from Kasper's(1992)definition of pragmatic transfer in the previous review together with the types of communication outlined in Table 2—1.

2.4.3.1　Pragmalinguistic failure

Pragmalinguistic failure occurs when the pragmatic force mapped onto a linguistic token or structure is systematically different from what is normally assigned to it by native speakers of the target language，or when conversational strategies are inappropriately transferred from the speaker's mother tongue to the target language (Thomas，1983：97).It may arise from two sources，i.e.teaching-induced errors and pragmalinguistic transfer—the inappropriate transfer of speech act strategies from one language to another，or the transferring from the mother tongue to the TL of utterances which are semantically/ syntactically equivalent，but which，because of different interpretive bias，tend to convey a different pragmatic force in the TL.

In intercultural communication，people often take it for granted that in the TL all the expressions can find equivalents in their mother tongue.If they take the two expressions in the two languages as semantically and syntactically equivalent，they often neglect their different pragmatic meanings.Pragmalinguistic failure concerns the pragmatic issues of the language itself.It can be perceived at two layers.The first-layer pragmalinguistic failure is caused by the facts that，because the speaker uses words inappropriately，and even uses them ambiguously，the hearer misunderstands the speaker's utterance meaning conveyed under certain conditions and the range it referred to.The second-layer pragmalinguistic failure occurs when the hearer either misunderstands the illocutionary force that the speaker wants to express or the speaker fails to express the illocutionary force explicitly in his utterance.The speaker thinks that what he himself says can be fully understood by the hearer，but on the contrary，the hearer misunderstands what the speaker means and makes a wrong pragmatic deduction(He，1988：228-9).

2.4.3.2　Sociopragmatic failure

Sociopragmatic failure mainly stems from sociopragmatic mismatches.It is described as a lack of sociocultural awareness or ignorance of“social conditions placed on language in use”or as stemming from“cross-culturally different perceptions of what constitutes appropriate linguistic behavior”(Thomas，1983：99).It occurs when nonnative speakers fail to choose the appropriate language form owing to a lack of the knowledge of cultural differences.Dif-ferent cultures tend to have very different ways of thinking，rules of speaking，social values and relative weights of pragmatic principles.The diversity in social value systems and the ways in which ideas are expressed in intercultural communication are not always well interpreted.Such being the case，it is inevitable to misunderstand or hard to cope with one's intention，since it involves the speaker's cultural knowledge of the TL，or the conflicts of the source culture and the target culture.For example，people living in English-speaking culture tend to verbalize their gratitude and compliments more often than native Chinese speakers and they tend to accept thanks and compliments more readily than the Chinese.Just take the following dialogue as an example：

(A young Chinese lady(CL)in the U.S.was complimented by an American lady(AL)for her lovely dress.)

AL：It's exquisite.The colors are so beautiful！

CL：Oh，it's just an ordinary dress that I bought in China.

Upon hearing the compliment，the Chinese lady was pleased but somewhat embarrassed.Following the typical Chinese fashion，her reply is a typical sociopragmatic failure，since the words of the Chinese literally convey a message quite different from what was intended.In the case of the Chinese woman，her response would have meant that the complimenter does not know what indeed a good dress is，otherwise，how could she get so excited about an ordinary dress？The implicature is that the American woman's taste in clothing was questionable(Deng & Liu，1989：38).

We have thus far examined two types of pragmatic failures.However，the distinction between pragmalinguistic and sociopragmatic failure is not clear-cut in some cases.The reason lies in that in different contexts，communicators may have different intentions and different understandings about the same utterance，or that form and function is not a matter of absolute one-to-one match.So an inappropriate utterance may be viewed as a pragmalinguistic failure from one point and a sociopragmatic failure from another.For example，in a department store if a Chinese salesclerk greets his English customers by“What do you want？”，which is often used in the Chinese context to greet customers，the English customers will feel s/he is very rude，because“what do you want？”is not an equivalent of the English customer-greeting expression“Can I help you？”or“What can I do for you？”So the sales clerk's utterance fails both pragmalinguistically and sociopragmatically(He & Yan，1986：53).

Generally speaking，pragmalinguistic failure is more languagespecific，while sociopragmatic failure is more culture-specific because it is related to“people's view of value，social distance，relative rights and obligation”(Hu，1994：29).Thomas claimed that“That which is related to cross-cultural failure is referred to as pragmalinguistic failure whereas that which has a non-cultural basis due to the social relationships and positions between individuals is referred to as sociolinguistic failure”(1983：99).No matter what they are，pragmatic failures may cause intercultural communication breakdowns.A more detailed discussion will be made in Chapter 4.

2.4.3.3　Pragmatic behavioral failure

Verbal communication goes hand in hand with nonverbal communication in human communication.They are inseparable and complementary to each other.Samovar et al.pointed out that most research experts think that in face-to-face communication only about 35% of the social content of the message is conveyed by verbal behavior，and the other messages are conveyed through nonverbal behavior.In fact，people speak not only with vocal organs，but also with the entire body(1981：155).Taking intercultural communication and FL teaching and learning as the point of departure，Bi Jiwan divided nonverbal communication into four types，viz.(1)body language，which involves communicative messages conveyed by basic postures，gestures and etiquette behaviors such as handshaking，kissing，hugging，smiling，touching and lady-first actions，and the movements of body parts such as head movement，facial expressions，eye contact，hand and leg movements，etc.；(2)paralanguage，which comprises silence，turn-taking，and various“nonsense”voices or backchannellings；(3)object lan-guage，which is made up of the communication messages provided by such aspects as the adornment of skin，the covering up of body odor，clothing and dressing artifacts，furniture and car；and(4)environmental language，which includes spatial messages such as kinesics，proxemics，notion of territory，space orientation and seats arrangement，temporal message，architectural design，home decoration，sound，lighting，color and signs，etc.(1999a：6 -7).

Nonverbal behaviors play a very important role in social interactions.When the verbal action is symbolically impossible or unsuitable in communication，interlocutors will strategically resort to nonverbal behaviors to compensate for it.Suppose that you go to the railway station to see your friend off.As the train moves farther and farther away，you and your friend cannot hear each other clearly.In this case，it is unsuitable for you to communicate by means of verbal behaviors；the best way for you to use is to wave your hand to bid farewell and to express wishes.Lv & Chen (1997：207 -209)noted that nonverbal behaviors are innate and universal in use，but its use is bound with the context of situation and the context of culture.The universal aspects of nonverbal behaviors are very limited for people in the world.They are people's facial expressions，firing a gun salute，flying a flag at half-mast and showing a yellow card and so forth.These nonverbal behaviors can be understood almost by people all over the world.Yet，in most cases，the meanings of nonverbal behaviors are not universal but vary from culture to culture.

There are many cases in point.For example，nodding to someone shows approval，agreement，and affirmation，which is equal to“是的”in Chinese culture or“yes”in English culture.But in some other countries such as India and Greece，it means the opposite.Shaking hands is a very common way to greet and say hello in many countries，but in New Zealand，people greet each other by means of touching each other's nose.In accordance with Hu Wenzhong's research，for Chinese students，intercultural pragmatic failure is mainly reflected in“cultural mistakes”，which he defined as“inappropriate or unacceptable language or behaviors for most native-English speakers”(1994：127).Therefore，different cultural mistakes have become the main factors that hinder Chinese students' communication with native English speakers in intercultural communication.Thus，it is of great importance for both instructors and learners to pay attention to nonverbal behaviors in ILP development and intercultural communication.And knowledge of nonverbal behaviors should and must be one of the components of communicative or intercultural competence.Only by conducting miscommunication research，contrastive pragmatics research and interlanguage pragmatic research，can we get an overall picture of pragmatic failure.