THE INTERFERENCE OF CANTONESE IN ENGLISH

Research Project I suggested clearly that first language inter­ference was a significant problem for Hong Kong learners of English.

We have investigated elsewhere the status“ of errors made by Cantonese speakers of Englishand their typical characteristics“ collated by e.g. Bunton (1989 and 1991), Potter (1992) and the Boyles (1991).

This section is devoted to an examination of more structural and psycholinguistic factors in interference.  Mere error collation with the occasional explanation does little to alleviate the sometimes extraordinary difficulty encountered by Cantonese learners of English in avoiding the direct transfer of structures of Cantonese into English language use.  The problem is actually more complicated than mere ‘imitation’ would explain. Moreover, as interference is identified by students as an important factor in their general difficulty in learning English in Hong Kong, it requires an investigation from the learner’s viewpoint (rather than a teacher’s or examiner’s viewpoint which is the one usually provided by a collation of typical errors).

Webster, Ward and Craig (1987) present a large body of data rather than a comprehensive theory of interference of the Sino-Tibetan Cantonese in the use of Indo-Germanic English.  In its practicality and attention to detail, the analysis is supremely useful for our purposes of following up the indications  of students  regarding  interference in Research  Project  I.  The analysis is broken down into four sections: morphology, syntax, lexis and ‘unclassified’.

Their numerous and detailed findings may be summarised as follows:

Suffixes -er, -est and -ing show no sign of interference but - ed, -s, -es and ‘s present difficulty for Cantonese learners.  This does not appear to have a first language cause. The cause is also not conceptual as Cantonese does have a concept of plurality (e.g. ngodeih, we from ngo, I). Students do not confuse he is/they are nor do they have problems with the morphological contrast go/went. The problem is said to be phonological (no final consonant clusters in Cantonese).

At the syntactical level, differing Cantonese structures: haih gau/it’s a dog - sihk gau/eat dog - gau moh yik/dogs don\t have wings - leads to the common omission of the article by Cantonese speakers of English. The structure di+noun (e.g. di gau hai bindouh/where are the (known) dogs combines with the aforementioned to produce the use of ‘the’ where English does not require it: I go to the bed early.  The form haih jek gau/it’s a dog (not two dogs) or /it’s the dog we know leads to the use of ‘the’ for ‘a’: I have the dog.  The form yat jek gau/one or a dog lead to the use of ‘one’ for ‘a’: suddenly one cat jumped into my garden. The form in Cantonese demonstrative+di+noun (e.g. nidi gau/these dogs) leads to the use of ‘those’ where English requires no article or ‘the’.

There are numerous ways in which Cantonese structures interfere in English at the phrase level, in noun phrases and relative structures. The three relative structures in Cantonese are performed:

1)   by simple juxtaposition: literally ‘Hong Kong is one+cl            small island have very big population’

2)   by placing the relative structure in front of noun with suffix -ge: literally ‘Hong Kong island is one+cl have very big population+adj small island’

3)   by placing the relative structure in front of the noun cluster immmediately preceded by go-/that and a relevant classifier: literally ‘I just buy that+cl book’.  

These structures are reflected in common errors such as:

1)   I saw a big car only carried one person.

2)   Better than those buses to Shek O.

Adjective modifiers hou/very and gei/quite are no problem in affirmative sentences but in negative sentences, gei is still considered to be ‘quite’ although in Cantonese it can mean ‘very’ leading to confusion: ‘It’s not quite easy to sell this product.’

Cantonese has a large number of words which can function as adjectives and as a corresponding noun:

jihyauh/free or freedom

Cantonese speakers frequently use the noun instead of the adjective (and less frequently vice versa):

I have freedom times (free time).

I want to be more confidence in speaking English.

In the area of word order, the Cantonese structure: literally, ‘is one so good man’ leads to the frequent use of ‘a such’ instead of ‘such a’. The Cantonese structure: literally, ‘we must build flyovers, the more man the more good’ is reflected in the error: ‘we must build flyovers as many as possible.’ The Cantonese structure: literally, ‘the street not enough wide’ is reflected in ‘ The streets are not enough wide.’

English and Cantonese have ‘contrastive systems out of synchronisation’. The Cantonese verbal system is formally much simpler, has differing ranges of contrasts not related to English usage and emphasises ‘state of the action’ rather then time.

The paucity of contrasts (unmarked, -jo form, -gwo form, contemporaneous, end of action and total action) in Cantonese make for little interference from them in English. The complexity of the English verb system produces a great number of errors.

Sometimes, however, certain English morphs (-ed) are equated with Cantonese (-jo form) producing: ‘When we arrived, it is raining ‘ equivalent to the Cantonese form: literally, ‘we

arrive+jo that time, fall rain’. -ing is equated to the contemporaneous form in Cantonese and causes difficulty for students not understanding its function as a verbal noun. The total action form in Cantonese is imitated by the use of ‘finished’ (tickets are finished).

In verb systems there is also some confusion between ‘will’ and ‘be likely’ explained by the Cantonese form ‘wuih’ being understood as the equivalent of ‘will’ leading to sentences such as ‘will he come tomorrow’ (intention: ‘is he likely to come tomorrow’), ‘I will make some dresses’ (intention: ‘I will probably make dresses from time to time’) and ‘It will often rain there’ (intention: ‘it often rains there’).

The English verb+modifier causes some confusion because of

Adverbials of time lead to confusion over word order because the normal unmarked position (e.g. in English, ‘we went shopping last Saturday’) in Cantonese is after the subject and before the verb: literally, ‘We last Wednesday go to buy things.’  This adverbial structure extends to people: literally, ‘I with Ah Bou go drink tea.’  The use of the pause marker ‘ne’ to emphasise the subject leads to the common imitation: ‘My little sister she is only three years old.’

The subordinating conjunction structure yan waih... soyih (because) in Cantonese leads to its imitation in English as:

Because...so. A similar interference problem is presented by the structure seuiyin...daahnhaih (although) and presents as:

Although...but.

Cantonese has no special structure for direct speech, using instead: literally, you know not know, he is who+interrogative particle/Do you know who he is?. The literal translation: you say give me hear, you like what+interrogative particle/Let me know what you like is a similar structure leading to errors: Do you know who is he, Please let me know what do you like etc..

The question tag haimhaih (literally, is not is) leads to overuse and incorrect use of ‘isn’t it?’ e.g. ‘We’ll come on Friday, isn’t it?’

Yes/No responses prove difficult for Cantonese speakers because a question may be answered by repeating the verb of the question (literally,: you have money? Have. Have not.) or by using  haih/mhai (is/is not). The latter response merely indicates agreement or disagreement with the questioner’s words and do not necessarily answer the question. Cantonese code of yes and no leads to the responses: You don’t like swimming? Yes (= no, I don’t like swimming).

Written Chinese has directives at the beginning of paragraphs which have no direct equivalent in English. These are: yihche (dictionary translation: moreover, in addition) and lihng yat fongmihn (on the other hand, besides). Imitation of these Chinese structures leads to overuse and improper use of moreover, besides etc..

Lexical items cause a great deal of interference:

tai = see, read, watch (TV)

teng = listen, understand speech, answer telephone

jyuh = stay, live (in a place)

muhn - bored, boring

yauh = have, there is

msyuhfuhk = uncomfortable, ill

waan = play (a game), have fun (leading to ‘he likes playing his bicycle’ meaning he likes playing around with his bicycle.

hoi = open, turn on (switch)

shan = close, turn off

jaahm = railway station, bus stop

yahn = people, person

keuih = he or she

sung = accompany, send, see off

bei = give, cause/get (causative sense), send a message

lihnjaahp = take exercise, practice (e.g. the piano)

yuhng = use, take (time)

je = borrow, lend

jeyuhng = borrow, use (e.g. ‘may I borrow your toilet’)

jeunbeihhou = ready, get ready (prepare)

nidouh/godouh = this place/that place (here/there) leading to ‘the scenery of there is very beautiful’

wan = look for, wandou = find (‘I found my student card missing.  find it in every railway station’).

dou = until, by the time that

mahn = ask a question, yikauhn = ask for something leading to:‘you may ask for him for his book’.

Lexical items combinig several words present difficulties:

heui Oman = got to Macau (no preposition in Cantonese)

but analogy with ‘to Macau’ leads to ‘go to shopping’.

cho feigei (literally: sit plane) = go by plane, fly leading to ‘sit bus’ (go by bus).

A number of inappropriate responses are caused by interference from Cantonese:

donyihm la!/of course, said emphatically the message is: ‘you were foolish to ask’.

mganyiu la!/never mind, often used instead of ‘don’t mention it’.

ngoh haih/ lit. ‘I am’ = ‘this is’ on telephone.

haih neihgo dihnwa, literally ‘it’s your phone’ / the call is for you.

Set phrases are commonly translated literally:

dimjouh/how do? i.e. ‘how do you do it?’ leads to sentences like ‘how to spell people?’ and ‘how to defend?’.

funyihnh neih leihsun/literally, ‘welcome your coming letter’ leads to ‘I very welcome your letter come’.

Other first language interference is not clearly classifiable:

‘one week I try to play once’ is a reflection of Cantonese structure, literally: ‘I try one week play one time’.

‘ I saw him two days before’ is a reflection of a Cantonese structure, literally: ‘I two day before saw him’.

Cantonese ‘dou’ (both,all) in negative structures leads to:

‘Buses and trams are all not enough’. ‘keuih leuhnggodou

mngaam’ lit. ‘two+classifer dou not correct’ leads to ‘both of

them were not right’.

‘Two or three’ is expressed as, literally ‘twothree’ (saamseibun) leading to: ‘I read three four books during the vacation’ etc..

Imitation of Cantonese structures frequently leads to a double subject:

I stomach hungry (ngoh touhngo/lit. I stomach hungry)

you eye red (neih ngaaanhuhng/lit. you eye red - you’re jealous.)

his body is very strong (keuih santai hou keung - his body very strong)

Webster and Lam (1991) clarifies and extends some areas of research from the above.

English spelling for Cantonese students presents difficulty because new words are regarded like new ideograms. My personal experience of teaching in an East Kowloon secondary school confirms this. Letters may be transposed just as strokes go askew in Chinese writing: from/form, calm/clam, ture/true, bady/baby or, less explicably perhaps: expamle/example, forigen/foreign, condifence/confidence.

‘When’ and ‘if’ are frequently confused as are degrees of certainty in general, e.g. ‘When there’s a fire, don’t use the lift.’  This is a result of the difference in meaning of yuhgwo/if, dangjahn and yatjahn/ if all refer to things which are likely to happen in the near future. Dong/when is used when it is in the more distance future and may be closer in meaning to if. All the Cantonese equivalents are not as positive as the English when. Fundamentally speaking, of course, the hypothetical and subjunctive functions of European languages are hardly reflected in Cantonese.

The complexity of the English verb system makes the correct use of suggestion difficult for Cantonese speakers:

She suggest I buy new car/ Keuih taiyih ngo maaih ga san che.I suggest go see movie/ Ngoh taiyih heui taihei.

Go swimming?/ heui yauhseui lo! or heuimheui yauhseui a?

The incorrect English is a direct translation of the Cantonese.

‘Let’s to go swimming’ and ‘He made me to copy out my homework’ arise likewise from the complexity of the English verb system.

Certain difficulties arise from differing concepts of certain activities, e.g. help, change, send.

‘Help’ is conceived in Cantonese as giving assistance by doing something with or for somebody, not as in English doing something with somebody or in a general sense ‘aiding’ someone.  Therefore Cantonese speakers may write ‘The maid helped us to cook dinner’ when ‘The maid cooked dinner for us’ is meant. 

In Cantonese, ‘change’ focuses on the new object: literally, ‘The bulb is broken. I’ll change a new one for it.’ (Ngoh wuih wuhn go san-ge dangdaam).

‘Send’ has a different semantic spread. The Cantonese equivalent does not differentiate between whether the sender accompanies the person/object or not: thus, ‘When you leave, I will send you to the airport’. The ‘inward-looking’ use of Cantonese, which relates everything to the subject of the sentence, may explain a most common Cantonese error: ‘He’s not come back yet.’ the boss’s office is regarded as the centre of his life and where he should normally be. This is an interesting interpretation.

There is a great deal of redundancy in Cantonese speakers’ use of English, e.g. ‘he returned back home’ (Cantonese equivalent has ‘back’/faan plus a particle but no verb so a verb is inserted in the English and ‘back’ is retained); ‘according to my opinion’ (Cantonese has redundant expression ‘gangeui’ which is reproduced in the English equivalent); and ‘the reason is because’ (yuhnyan haih yanwaih’ has redundancy when translated literally).

Because Cantonese uses fewer prepositions, in general, over-use of prepositions occurs to compensate for this perceived lack when transmitting ideas formed in Cantonese: ‘to discuss about’, ‘I went to shopping’ etc..

Barbara Chan (1991) investigated secondary form six students’ errors with passive and transitive constructions. Transitive verbs exist in Cantonese. Negative transfer occurs ‘with compound sentences involving more complicated structures’.  Objects occurring in the first clause of complex sentences in English are repeated in the second clause after verbs, but not in Cantonese. The passive is also not strange to Cantonese speakers.

Its occurrence is however quite rare, one explanation being that Chinese is a ‘topic-prominent’ language rather than a ‘subject-prominent’ language like English. The errors of Cantonese speakers with the English passive are said to be of three main types: Inappropriate use, failure to use passive where appropriate and errors in the passive construction. The first type of error is said to be caused by ignorance of rule restrictions and negative transfer; the second is caused by the fact that Chinese and English operate on different rules and by negative transfer; the third is caused by incomplete application of rules and mixing up of forms.

Sung (1991) identifies English as a subject-prominent language rather than a topic-prominent language like Cantonese.  Cantonese learners thus have tendencies towards: putting the topic in initial position, avoiding the passive and ‘dummy subjects (there is, there are), using double subjects, omitting coreferential constituents (e.g. any) and putting the verb in final position.

These tendencies feature heavily in errors and should be identified in English language teaching as sources of interference.

BEYOND ERROR CORRECTION AND INTERFERENCE ANALYSIS

More important perhaps than a mere cataloguing of errors (a first level) and the fundamental sources of interference (a second level) is some speculation regarding the reasons why:

i)    students are prone to interference

ii)   what efforts are made by teachers to overcome interference

This may be termed the ‘third level’ of interference study.

My Hong Kong students often appeared to be under the impression that English was an‘equivalent’ sign system which mirrors Cantonese in its internal logic and relationships. Moreover, students have the impression that performing English is a matter of relating English to Cantonese, not relating English to other English elements in discourse. The idea that English was related only to other parts of English was a separate and autonomous sign system was novel, even revolutionary, for my students. It may well be that students are never taught to see English in these terms and that interference is encouraged, perhaps even created, by teaching methods which, for example, operate by rapid code-switching (see R.K Johnson 1983) and mix (see Gibbons 1987).

Another fundamental problem in interference is that students may not have a true sense of the validity of their own language as an autonomous and logical sign system. S. J. Matthews (1992) has pointed to the myths commonly believed by students:

‘Cantonese has no grammar’ and ‘Chinese grammar is the rules and conventions of written Chinese’. These are reinforced by ‘informed’ linguist beliefs such as: ‘Chinese grammar, especially morphology is simple’, ‘Chinese grammar has developed to the level of simplicity’ and ‘There are languages, such as Chinese  and Siamese, that make no grammatical use of  elements that  do not at the same time possess an independent  value  as radical elements.’

SOME PSYCHOLOGICAL, DIDACTIC AND SOCIOLOGICAL FACTORS IN

INTERFERENCE: ATTITUDE TO ENGLISH, STATUS OF ENGLISH, TOLERANCE

TO ERROR AND ERROR INDUCED VIA TEXTBOOKS.

Interference may be promoted by a number of psychological conditions such as low esteem for the target language, tolerance to error in the community and the induction of errors by inaccurate teaching material.

Studies (see Fu (1975); Lyczak, Fu and Lee (1976); Pierson, Fu and Lee (1980); Bond and Lee 1980; Pierson and Bond (1981);

Young Pierson and Giles (1986); Pierson (1987)) tend to show that negative attitudes“ towards English do not figure significantly in the picture of English language learning difficulty in Hong Kong in that they do not hinder secondary students’ performance. Students accept English as the medium of instruction but it may be that students form improper ethnolinguistic attitudes which may hinder their ultimate“ performance. English is constantly associated in the local students’ psyche with intelligence and power, as something desirable but suspect and even arrogant, like foreigners are regarded in general (see Fu 1975 in particular). The language is associated with power and prestige rather than spontaneity and friendship. It is, as I have referred to it elsewhere, a Parental language. The weakness of the approach of Pierson, Fu at al. is in their data collecting technique. This is nearly always aimed at one response rather than a conflicting variety of responses. This problem is gone into in Project I. The rationale for my investigations was some thinking on transferring the conceptual framework of Transactional Analysis to the question of attitude and approach to study (Adams 1991). I devised the following table to show how ego states may be employed to illuminate ‘attitude’ and ‘approach

’The studies mentioned above nearly always aimed questions at the Parent of the subjects to elicit an ideal response. My project aimed to allow for deviant responses and conflicting responses. It gives, I believe, a clearer picture of the inherent confusion in the minds of local students learning English in Hong Kong.

Stead (1979) makes the assertion that there is a ‘progressive tolerance to error’ in Hong Kong classrooms, a view which is partially supported by some other studies (e.g. Hou 1983 and Pearson 1977).  A recent study by Green (1991) investigated teacher perceptions of the relative gravity of errors in written English in Hong Kong and found that non-native teachers marked more harshly than native speakers. Verb-related errors were judged to be the most serious by both native and non-native teacher. Preposition errors and the definite article revealed similar grading.  Pluralisation was ranked seventh by non-natives but fourth by natives; spelling was ranked fifth by non-natives and eighth (last) by natives. Native speakers also differentiated much more in the scores given to categories of error.

Text books produced in Hong Kong sometimes serve to perpetuate errors or ‘non-standardisms’ by reinforcing patterns of error which already exist in student usage. Newbrook (1991) argues that the lack of native-like intuitions about the English language amongst secondary school teachers in Hong Kong may well fail to dispel text-book errors and, combined with their own errors, makes for a bad example. Certain errors become ‘institutionalised’. Moreover, according to Newbrook:

‘It is not unknown for Hong Kong people who have acquired English to a level markedly higher than that prevailing in the community at large to resist further correction of their remaining localisms on the part of native speakers or others with a still more secure command of English. This attitude is sometimes found even amongst students, notably at tertiary level; students may simply refuse to believe that forms which they have been using since the early stages of learning English, and which their teachers have taught and endorsed, are non-standard or heard only in Hong Kong.’ (Newbrook 1991:24)

There is a wide range of ‘little helper’ textbooks available to students in Hong Kong and many are simply inaccurate. Students appear to have faith in them, perhaps duped by the show of qualifications of the writers. They sell well. The books are frequently badly proof-read but a more serious problem is caused by ‘non-careless’ errors (institutionalised and fossi­lised errors). Newbrook quotes examples of books which have one error every three pages and another (very widely used) book where one in three entries is erroneous or doubtful.  Another problem arises in books which, although generally accurate with regards to what constitutes correct usage, give incorrect salience to certain errors at the expense of others, argues Newbrook.  Other books simply give translations without anyrationale behind them.

Hepburn (1991) questions the ability of error analysis to predict language learner errors to any significant extent. Derived from contrastive analysis, error analysis does not account for the significant correlation of errors made by learners all over the world with different mother tongues. The way English is taught provides a more accurate means of under­standing how errors arise.

In  primary school textbooks in Hong Kong, the ordinary use  of English  is  misrepresented  in a number of ways,  e.g.  in  an attempt to teach the present continuous, sentences like ‘He  is sitting  down’ are taught as examples thus leading students  to produce  erroneous sentences by analogy: ‘I am seeing  the  bus coming’.

The classroom does not present a natural environment for teach­ing the present progressive. In addition, the order of teaching language items leads students into a great deal of confusion: the simple present -s is taught in one book together with the reply  ‘No, he does not’ and ‘Does he live?’ The -s is per­ceived as being redundant. This situation could be remedied if ‘formal features associated with different forms and functions are presented and practiced in separate, distinctive and authentic contexts’ (Hepburn 1991:66).  This is rarely the case in teaching texts in Hong Kong.  An additional difficulty is provided by the practice of text book writers to present con­trasting items simultaneously (e.g. I ate the apple/I’ve eaten the apple). This leads more often to confusion than to learn­ing.  Frequently also, tried and tested means of teaching in European contexts e.g. ‘this is a book’ leads to confusion as the language needs of Cantonese speakers are not taken into account (for whom the indefinite article and the unfamiliar sounds of th-, z and k may prove distracting).

CONCLUSION

Interference is a more complicated issue than mere contrastive analysis or error analysis would suggest. Intractable and ‘institutionalized’ errors abound in Hong Kong and are both evidence of and a causative factor in English language learning difficulty. The power of the first language’s structures to disrupt the acquisition of accurate competence in English is not exactly quantifiable but appears to be potent for a number of reasons:

a)   the failure of teachers to stress the autonomous nature of English and to bring about awareness of mother tongue grammaticiality;

b)   the inaccuracy and unsuitability of certain teaching materials;

c)   negative attitudes towards the cultures of the target  language brought about, partly, by colonial sensibility which lowers student and teacher sensitivity to error;

d)   the pervasiveness of Cantonese culture which makes contact with target cultures and speakers difficult and socially unusual.

e)   negative attitudes towards academic excellence brought about by exaggerated credentialism (see section 8).

These reasons, which are mostly hypothetical, do not really account for the apparent resistance of students and of the teaching profession as a whole in Hong Kong to renewal and change (see Morris 1992). Such a passive collective conspiracy if it exists - is difficult to explain even speculatively. Investigation of such a collective societal conspiracy - or game - would perhaps yield more fruit than yet another book of ‘typical Hong Kong errors’.

Following this analysis of the main area of difficulty identified in the first research project, we proceed by looking at another important area: the teaching and learning situation in Hong Kong which we have termed the ‘sociological’ perspective.