Chapter One Introduction
1.1Research background
Since the early 1990s，the rapid development of corpus linguistics hasrevolutionized nearly all research fields where language plays a major role，andTranslation Studies is no exception. The availability of large-scale corpora of bothsource and translated texts and the rapid development of corpus-based approacheshave brought new perspectives and more vigor into Descriptive Translation Studies(Toury，1995)，which is mainly centered on translation as a product and exploresfeatures of translated texts per se，aiming to answer the question of "why a translatortranslates in this way” instead of "how to translate" (Holmes, 1972/1988). One of themain strands of corpus-based Descriptive Translation Studies involves examining theso-called translation universal hypotheses, which are concerned with commonlinguistic features of translational language as a "third code" (Frawley, 1984) differentfrom both the source and target languages. According to Baker (1993: 243)，translation universals refer to "features which typically occur in translated texts ratherthan original utterances and which are not the result of interference from specificlinguistic systems?” They are recurrent common features of all translated texts whichare distinct not only from their source texts but also from non-translated native textsin the target language. Moreover, these universal features are thought to beindependent of the influence of the specific language pairs involved in the translationprocess.
……….
1.2Research justification
As mentioned above, although explicitation has been widely investigated,previous studies have been criticized for substantial problems such as the debatablemonolingual comparable corpus-based model, the lack of adequate comparability ofthe corpora used，and the restriction to specific genres, which may invalidate thefindings obtained.As corpus-based Descriptive Translation Studies is mdnly centered on translationas a product，until recently explicitation has mainly been explored on the basis ofmonolingual comparable corpora of translated texts and non-translated texts in thetarget language, excluding source texts. However，as Kenny (2005:157) argues, thesource text must be "an important conditioning factor" in the production of anytranslations, and as such translation universal studies should combine comparablecorpus data with bilingual parallel corpus data which consists of the source texts andtheir corresponding translations.A number of prominent studies on explicitation (e.g. Olohan and Baker, 2000;Olohan, 2001; Baker, 2004) are based on the subcorpora of the pioneeringTranslational English Corpus (TEC) and the fictional components of the nativecorpus British National Corpus (BNC). However, these two subcorpora of translatedand non-translated English texts, in a strict sense，are not perfectly comparable. Baker(2004) also concedes the imbalance between the texts in the two subcorpora in termsof size and genres.
………
Chapter Two Literature Review
2.1 Explicitation
The notion of "explicitation" was first put forward by Vinay and Darbelnet (1958)as a translation technique, which refers to "the process of introducing information intothe target language which is present only implicitly in the source language, but whichcan be derived from the context or the situation" (ibid.:8, as cited in Klaudy, 1998: 80).Vanderauwera (1985) identifies a variety of explicitation techniques adopted bytranslators, such as using supplementary explanatory phrases, resolving source textambiguities，and making a heavier use of repetitions. Then explicitation as ahypothesized universal feature of translations is formulated by Blum-Kulka (1986:19)in what is regarded by many scholars as the first systematic study of explicitation:The process of interpretation performed by the translator on the sourcetext might lead to a TL text which is more redundant than the SL text.This redundancy can be expressed by a rise in the level of cohesiveexplicitness in the TL text. This argument can be stated as "theexplicitation hypothesis，，，which postulates an observed cohesiveexplicitness from SL to TL texts regardless of the increase traceable to thedifferences between the two linguistic and textual systems involved.
……….
2.2 Reformulation markers
Reformulation is a discourse function by which the speaker/writer restates, in thesubsequent linguistic unit，information provided in the preceding linguistic unit. Thetwo linguistic units of reformulation can be marked off by punctuation, e.g. commas,colons, dashes and parentheses. They can also be signaled explicitly by overt lexicaldevices such as namely, that is to say and in other words，which are commonlyreferred to as "reformulation makers" (e.g. Cuenca, 2003; Murillo, 2004; Blakemore,2007; Hyland, 2007). Reformulation markers are also known as “discourse markers ofreformulation”，"reformulators" and "apposition markers" (cf. Meyer, 1992;Blakemore, 1993; Mutesayire，2005; Del Saz Rubio, 2007)，which are a type ofdiscourse markers and a type of connectives claimed by some scholars (cf. Fraser2005).There are different views in the literature about what lexical expressions shouldbe admitted as reformulation markers. Cuenca (2003) observes that reformulation isgenerally based on an equivalence operation, since the two utterances are differentways to express the same idea (i.e. the second utterance serves as a paraphrase of thefirst one). However, she also argues that "reformulation is more than a strictparaphrase", but that “[i]t should be considered a complex semantic category thatranges from strict paraphrase to other values such as specification，explanation，summary or denomination，and even to non-paraphrastic meanings such as -implication, conclusion and contrast" (ibid.; 1072).
………….
Chapter Three Research Methodology...........23
3.1Corpora used in the present study..........23
3.1.1Comparable corpora..........23
3.1.2Parallel corpora..........24
3.2Tools used in the present study..........25
3.2.1WordSmith Tools 5.0..........25
3.2.2ParaConc..........26
3.2.3Log Likelihood Ratio Calculator..........26
3.3Data retrieval procedures..........26
4.1 Reformulation markers in comparable corpora..........32
4.2 Reformulation markers in parallel corpora..........40
Chapter Five Conclusion..........58
5.1Major findings of the present study..........58
5.2Implications of the present study..........60
5.3Limitations of the present study and suggestions..........61
Chapter Four Results and Discussions
4.1 Reformulation markers in comparable corpora
Table 4.1 presents the frequencies of commonly used reformulation markers inthe two comparable corpora of native and translational English. It can be seen that thetranslational English corpus COTE makes more frequent use of reformulation markersthan its native counterpart FLOB (536 vs. 380)，and the difference in the overallfrequencies in the two corpora is statistically significant as indicated by LogLikelihood (LL) tests. As reformulation makers area type of explicative devices，the substantially higher frequency of reformulationmarkers in translational than native English provides evidence in support of theexplicitation hypothesis.A closer examination of the detailed data in Table 4.1 suggests perceptiblyvarying frequencies of use across inpidual reformulation markers, ranging from oneinstance to 126 instances, while the majority lie somewhere between 20-40 instances.Although the overall frequencies of reformulation markers are significantly higher inCOTE than in FLOB, not all inpidual reformulation markers are more common inCOTE than in FLOB. As can be seen from the LL scores and p values in the rightcolumns of Table 4.1 only about half of the reformulation markers are morefrequently used in COTE than in FLOB，with five of them being significantly morefrequent, which are i.e. (abbreviation of Latin id est namely，that is, that is to say andor rather. In contrast, reformulation markers with "mean/means/meant" are invariablymore common in FLOB than in COTE, and the first three of them are significantlymore common as indicated by LL tests (/ mean/meant this means/meant and thatmeans/meant). It can also be noted that several reformulation markers are particularlyinfrequent in both corpora, such as viz (1 in FLOB and 2 in COTE), to be moreprecise/accurate (2 in FLOB and 1 in COTE)，to put it differently (1 in FLOB and 1 inCOTE), and or better still (1 in FLOB and 2 in COTE).
…………
Conclusion
The present study has explored explicitation in translational English via a casestudy of reformulation markers, with combination of comparable and parallel corpusapproaches and a combination of quantitative and qualitative analyses. It attempts tooffer more insights into the explicitation hypothesis, or more generally into translationuniversal hypotheses.On the basis of two comparable corpora of native and translational EnglishFLOB and COTE, it is found that translational English displays a significantly higherincidence of reformulation markers in comparison with comparable native English. Asreformulation makers are a type of explicative devices, the substantially morefrequent use of reformulation markers in translational than native English lends somesupport to the explicitation hypothesis. In particular, translational English tends to usemore literate forms of reformulation markers than its native counterpart.The comparable corpus analyses also reveal that the distribution of reformulationmarkers varies considerably across genres. While reformulation markers are morecommon in translational than native English in both fiction and non-fiction, nosignificant difference between the two varieties of English is found in theirfrequencies of reformulation markers in fiction，whereas the difference is significantin non-fiction. This means that non-fiction translation displays a higher degree ofexplicitation as compared with fiction translation.
………
References (omitted)