PreprintPDF Available

Contrastive linguistics: Approaches and methods

Preprints and early-stage research may not have been peer reviewed yet.


This paper reviews the various approaches and methods of contrastive linguistics.
Cairo University
The Department of English Language and Literature
PhD Program in Linguistics
Contrastive linguistics: Approaches and methods
By Mayada Tawfik Zaki
Table of Contents
I. Introduction---------------------------------------------------------------------- 3
II. Contrastive analysis hypothesis and transfer theory---------------------3
III. Theoretical and applied CA-------------------------------------------------- 4
IV. Methods of analysis in CA-----------------------------------------------------6
V. Microlinguistic analysis ------------------------------------------------------7
5.1 Grammatical level -----------------------------------------------------------8
5.1.1 Models of grammatical CA--------------------------------------9
5.2 Phonological level ----------------------------------------------------------9
5.2.1 Models of phonological CA -----------------------------------10
5.3 Lexicology level------------------------------------------------------------ 11
VI. Macrolinguistic analysis------------------------------------------------------11
VII. Arguments against CA------------------------------------------------------- 12
References ---------------------------------------------------------------------- 13
I. Introduction
The main purpose of contrastive analysis is to investigate the structural features
of languages and find the areas of difficulty in the learning of a second language. It
focuses on the intermediate stage between L1 and L2 rather than the target
language itself. Though the purpose seems more related to psychology, the only
method to reach a description of languages is linguistics. Hence, it is also called
contrastive linguistics. The purpose of this paper is to review the contrastive
analysis hypothesis, theoretical and applied CA, methods used to contrast language
subsystems; in addition to a special focus on the various microlinguistic levels and
models of analysis as well as on the macrolinguistic CA approach. Finally, some of
the arguments aroused against CA will be discussed in an attempt to illustrate the
limitations of CA in order to bridge these gaps in following CA research.
II. Contrastive analysis hypothesis and transfer theory
Contrastive research first started in the 1940s by Charles Fries. Then in
1957,Robert Lado developed the contrastive hypothesis stating that contrasting two
languages would help predict the features that would represent difficulty or ease in
learning a second language due to realizing the differences and similarities
between the first and the second language. Lado assumed that second language
learners transfer the forms, meanings and the way they are distributed in their
native language to the second language. In the preface of „Linguistics across
cultures (1957)‟, Lado stated that systematic comparison of languages can help
predict and describe the difficult features that can face a second language learner in
learning L2. This perspective relied on structuralism as stated by Bloomfield
(1933) which further assumed that the structure of any language is finite and hence
can be determined and compared to another language (Byung-gon, 1992).
The second perspective to guide the contrastive analysis hypothesis was the
school of behaviorism due to its assumption that the difficulty or easiness of
acquiring a second language is rendered to the already acquired habits of the first
language. This, therefore, led to the emergence of another theory called transfer
that mainly relied on the assumption of transfer of habits from the native language
to the learned one (Corder, 1971 in Byung-gon, 1992). The transfer theory
complements the contrastive analysis Hypothesis and serves its applied purpose. It
illustrates how certain linguistic features of the second language can be more
difficult to acquire than others. Stockwell et al. (1965) claimed that when the
features of the contrasted languages are similar, positive transfer occurs; while in
cases of different features, negative transfer shows, causing difficulty of
acquisition. And in cases of no relation between features, zero transfer takes place.
With regards to the prediction of difficulty level in second language acquisition,
CAH was claimed to imply strong, weak and moderate versions. The strong
version is related to the potentiality to predict the difficulties in second language
according to systematic analysis of language. The weak version is the need to
depend on the „best linguistic knowledge available‟ (Wardaugh, 1970) in order to
detect difficulties in the second language. The moderate version is refered to by
Brown stating that confusion may occur more frequently with similar features
rather than different or non-existing ones (1987).
III. Theoretical and applied CA
Fisiak (1990) in his article „on the present status of some metatheoritical and
theoretical issues distinguished between theoretical and applied contrastive
research, equivalence and tertium comparationis, linguistic theory and contrastive
linguistics as well as discussing the relationship between contrastive analysis and
In the first part, Fisiak (1990) argued for the importance of distinguishing
between theoretical and applied contrastive analysis research. The difference
mainly concerned the purpose of each. The purpose of theoretical research is to
develop models of language analysis, describe languages and explore similarities
and differences between languages; while applied CA research aims at examining
language for bilingual education, translation or any practical specific purposes;
thus, making the latter more relevant to „psych-sociolinguistic settings‟ (Sajavaara,
1985 in Fisiak, 1990). Such framework, hence, reinforces the involvement of
intrapersonal, interpersonal and organizational levels in the investigations. The
intrapersonal level focuses on cognitive decoding and encoding of messages in the
speaker-hearer. The interpersonal level is concerned with the use of language in
communication or in other words with discourse analysis and ethno-methodology
of speaking. The organizational level is more concerned with the constraints set on
a native or a foreign language in the society.
In the section on tertium comparationis, James (1980) examined the
effectiveness of surface structure, deep structure or the semantic level, and
translation which also included pragmatic level and concluded that the semanto-
pragmatic translation equivalence is the most effective criteria for contrasting L1
and L2. In contrast, Krzeszowski (1981) argued for the sentential equivalence
which includes semanto-syntactic level rather than pragmatic. He claimed that it is
the core syntactic contrastive analysis. A more general argument was proposed by
Janicki (1985) that both are valid according to the purpose of the analysis.
In his contrast of theoretical and applied CA, Fisiak claimed that theoretical
research focuses more on abstract concepts such as, “grammatical categories, rules,
functions, and constraints”, like for example using the structure of NPs or
constraint on WH-movement; while applied research focuses on psycholinguistic
elements that are more perceptual even if the contrasted variables are syntactically
incomparable at all. Applied CA research must propose hypotheses and solutions
for problems (Fisiak, 1990).
IV. Methods of analysis in CA
Whitman (1970 in Byung-gon, 1992) has mentioned four steps to analyze
languages. First, the researcher writes description of the two languages, second,
forms are selected from the two descriptions; then the two selected forms are
compared and finally features of difficulty are predicted. Comparison of the two
language subsystems should be through the same model of description.
Nevertheless, a dilemma still can occur as the model used can be of favor of one
language rather than the other. One proposed solution for this dilemma was to use
the translation theory whereby each language can be described by its favorite
model then translated into an artificial „etalon language‟ (Melchuk, 1963) that can
enhance the features of L1 and L2 constructions. Another solution was applying
description bias to the second language and how it is used by the second language
learner more than the focus on understanding L1.
Comparing the two language subsystems involves several steps: First, the
gathering of data of the system to be compared in the two languages. CA uses
translations of the two languages without worrying about the bias of different
meanings due to its focus on general rules or systems rather than the focus on the
translated meaning. CA aims at generalizing its findings on the grammatical
systems of compared languages. Second, description of the realizations of each
grammatical category in each of the two contrasted languages, such as, for
instance, determining the realization or the context of using the indefinite article in
English and Russian. Third step is the addition of new data with their translation to
the corpus and then modifying the rules to include the new data. Finally in step
four, a formulation of the found results of the contrasted data is determined either
in the form of equations or operations. The formulation was either in the form of a
set of instructions that can be applied to both language grammars (Harris, 1954 in
James, 1980) transfer rules or equations which differ from transfer rules in that
they do not show which language is being converted to the other and hence lack
the directionality of the transfer rules. Moreover, equational statements show the
phonological representations of the category which helps to reveal the variety of
forms for a specific category in contrast to transfer rules which focuses only on
structural or syntactic depiction.
V. Microlinguistic analysis
In their effort to reach a reliable contrast of two or more languages, CA
linguists set fixed linguistic categories to describe the different languages in an
attempt to have constant factors. On the microlinguistic level, the language
variables are organized according to three levels- phonology, grammar and lexis-
and categories- unit, structure, class and system. In the traditional approach of
analysis, the linguistic level was described separately without reference to other
levels, describing phonological features did not include any reference to
grammatical ones, for example. Then merging the description of different levels
was found later to be inevitable. In Hetzron (1972 cited in James, 1980)
homonymy which was given as a reason to support the syntactic order in Russian.
The principle of linguistic level is analyzed by CA to observe the shift from one
level to the other. For example, Russian questions are distinguished by their
intonations while English questions are formed by the fronting of verb do
syntactically. This is described as „a phonology-to-grammar level shift‟.
5.1 Grammatical level
In the pursuit of reaching fixed organizational framework for the description
of languages, Halliday (1961 in James, 1980) set four grammatical categories- unit,
structure, class and system- that he described as “universal, necessary and
sufficient” for describing any language. The unit category includes the sentence as
the biggest unit of analysis which is then followed by clause, phrase, word and
morpheme. From this perspective, CA therefore does not analyze more than the
sentence level. It may observe, for example, that the same sentence has different
number of clauses across the two languages. Structure is the second category and it
refers to the order of the components in the sentence structurally or that of sounds
in a word phonologically. In English for example the sentence is composed of
subject predicate, compliment, and adjunct and phonologically words can be cccvc
or vccv. The adjective in French occurs in a post-nominal position while in English
it is pre-nominal. The third category „Class‟ depends on the place a specific unit
may occupy in the sentence structure, eg. Any phrase that can occupy the adjunct
is considered one of the class of the „Adverbial phrase‟. The last category System
includes a variety of options for the same element that can occupy the same place
in the sentence, such as plural and singular nouns in English. In Arabic there is also
5.1.1 Models of grammatical CA
Contrasting languages requires using the same model of analysis because
each model focuses on certain features, and hence comparing features analyzed by
two different models will make it difficult for the linguist to determine whether it
is a trait of the data or the model. Using the same model contributes to having
constants and reliable CA data. A variety of models of analysis are used by
linguists, two of which are the structural or the Taxonomic model and the
Transformational generative grammar. Through the taxonomy model, structuralists
proposed the Immediate Constituent Analysis technique whereby any complex
grammatical structure is divided into two constituents AB+C or A+BC according
to which parts should be in order or can be omitted. The phrase rather nice girl can
have „nice girl‟ as one construction but „rather nice‟ cannot be accepted as one
construction. Such analysis does not account for meaning, it only considers
construction types „syntagmatic‟ and possible elements for each structural position
„paradigmatic‟. With the eminence of Chomsky‟s universal grammar, language is
analyzed by the Transformational generative grammar in which a difference
between surface structures and deep structures of the sentence is highly considered.
Deep structure is considered universal and hence allows only for contrasting
different surface structures across L1 and L2. Generative grammar focuses on the
intermediate structure where diversion across the two contrasted languages
appears. Other models of analysis have also influenced contrastive analysis
5.2 Phonological level
In acoustic phonetics, contrastive linguistics focuses on sounds that have
physical similarities between L1 and L2 and then tries to determine the differences.
Similar sounds in two languages can be of different functional importance. For
example two allophones in one language can be considered as two different
phonemes in the other. Contrasting two sound systems involve four steps. First, a
phonemic inventory of the two languages is drawn. Second, phonemes of the two
languages are equated. Third, the different phonemes and allophones are listed.
Then, the distributional restrictions or on the context of the phonemes and
allophones are determined for each language. For example, the sound [ŋ] in
English and Spanish in English it is an allophone of /n/ while in Spanish before /h/
and /w/ such as [estraŋ‟hero]. Phonemes can contrast in one of the following ways
(Politzer, 1972). Two similar phonemes in the contrasted languages does not mean
equation of their allophones since one phoneme can have allophones and the other
does not show allophones at all. Another contrasting feature shows when the same
sound is considered a phoneme in one language and an allophone in the other.
5.2.1 Models of phonological CA
Taxonomic phonology and generative phonology are the two main models
of analysis in phonological CA. The taxonomic model aims at stating the two
phonological systems of the two languages and the variations of similar sounds.
The phonemic approach indicates that errors of pronunciation by the L2 learner
occur because of phonemic asymmetries and allophonic variations which may lead
to a foreign accent. However, this taxonomical model fails to highlight the
difference between receptive and productive difficulty. Generative phonology the
other hand, depends on the concept of transformation of deep structures into
surface structures which are psychological non-realistic; thus, making the
taxonomic approach more practical.
5.3 Lexicology Level in CA
In Determinism language sets the structure of reality leading to different
view of reality by different language communities. One CA model depends on
word fields where the lexicon is grouped according to “semantic, cognitive,
attitudinal, or notional areas of concern.” For example, verbs are grouped in a
notional class of verbs that refer to speech acts such as „say, speak, tell, and talk‟ in
a study by (Lehmann, 1977) then they are to be compared to their equivalence in
German. One argument against the notional class is that it can never be objective
and does not have well set criteria for adding a word to a specific word field.
Another model of contrasting lexis is the semantic components. This approach
assumes universality of some components that exist in all languages and hence
creating a lexical inventory of features is considered possible, an assumption which
is criticized by the fact that each language may have its subset under the universal
features. In CA two approaches can be followed. L2 lexemes are specified via an
inventory then each lexeme is analyzed according to the Symantec components.
The second approach is the translation equivalence whereby words are translated
tentatively; then checked by components to confirm if they really are similar.
VI. Macrolinguistics
Viewing microlinguistic analysis as idealistic causing regularization and
decontextualization of data, focus has turned to analyzing bigger chunks of
language and how they are organized in texts on the one hand, and how language
functions in discourse as well as in its socio-cultural setting (Coulthard, 1977) on
the other. In summary, CA studies text either through textual characterization,
text type or translation of texts. In the first approach, textual characterization, data
collected according to the preference of specific features of textual cohesion in
each language. Thus, texts are observed in the two languages for the type,
frequency and context of cohesive devices. Wonderly (1968) found, for instance,
that the use of ellipses enhances style in English while repetition may be a
preference in other languages as the Mayan languages of Central America. Text
typology is another approach in macrolinguistic contrastive analysis. It compares
types of text that have the same function in the two languages, such as comparing
rituals or reports. The third approach is translated texts which are criticized for
their potential to being distorted by formulations of the source language.
Contrastive analysis of discourse and pragmatics is beneficial for providing the
second language learner with how to interact in the community and context of the
second language. It includes comparative analysis of conversation.
VII. Arguments against CA
One of the criticisms on structuralism CA is that it is contrasting parts or
subsystems of languages and not the two languages in their entirety. Lee (1968)
argued that language parts are interdependent and “determinative”. Lee thought it
is “naive” to aim at only the parts that can hinder L2 acquisition at the beginning
stages. However, James (1980) claimed that languages are complex systems that
need to be divided into smaller subsystems for analysis purposes.
Wardhaugh (1970) argued that a strong feature of CAH is that relevant to
comparing two language systems to predict the challenges in learning the second
language is applicable. However, a weak point of CAH is that it is difficult to
investigate only the best linguistic data provided in order to explore language
difficulties for second language learners.
One of the arguments against CAH is that it did not account for the different
types of second language learning such as natural and mediated or second versus
third language. Moreover, it did not realize the age of the second language learner
(Byung-gon, 1992).
The emergence of transformational and generative grammar and the belief in
the innateness of language acquisition was the greatest challenge to CAH. After
Chomsky (1964), it was claimed that all natural languages have common features
and hence the language learner already knows a lot about the second language; in
addition, the deep structures of the two compared languages are the same and
hence differences are only in surface structures.
Further arguments also considered the results of empirical research that
found mistakes which were never predicted by contrastive studies (Lance, 1969).
In contrast, many of the predicted mistakes by CA were not seen by second
language learners as well as the interference of factors other than the transfer of
L1. However, CA is beneficial for translation studies, the study of bilingualism and
second language teaching.
Bloomfield, L. (1933). Language. NewYork: Holt, Rinehart and Winston
Brown, H. D. (1987). Principles of language learning and teaching. 2nd ed.
New Jersy: Prentice Hall.
Byung-gon Y. (1992). A review of the contrastive analysis hypothesis.
Dongeui 19, 133-149.
Chomsky, N. (1964). Current issues in linguistic theory. Mouton: The Hague
Corder, S. P. (1971). Idiosyncratic dialects and error analysis. IRAL (5).
Coulthard, M. (1977). An introduction to discourse analysis. Longman
Fisiak, J. (1990). On the present status of metatheoritical and theoretical
issues in contrastive linguistics. In Fisiak, J. (ed.), 3-22.
Fries, C. (1945). Teaching and learning English as a foreign language. Ann
Arbour: University of Michigan Press
Halliday, M. (1961). Categories of the theories of grammar. Cited in James,
C. (1980). Contrastive Analysis. London: Longman.
Harris, Z. S. (1954). Transfer grammar. IJAL, 20 (4), 259-270.
Hetzron, R. (1972). Phonology in syntax cited in James, C. (1980).
Contrastive Analysis. London: Longman.
James, C. (1980). Contrastive Analysis. London: Longman.
Janicki, K. (1985). On the tenability of the notion pragmatic equivalence in
contrastive analysis. PSiCL 20, 19-26.
Krzeszowski, P. T. (1981). The problem of equivalence revisited. IRAL 19,
Lado, R. (1957). Linguistics across cultures. Ann Arbor, Michigan:
University of MichiganPress.
Lance, D. M. (1969). A brief study of Spanish English bilingualism. College
Station: Texas: A&M University.
Lee, W. R. (1968). Thoughts on contrastive linguistics in the context of
language teaching. Georgetown monograph 21, Washington, D.C.:
Georgetown University Press.
Lehmann, D. (1977). A confrontation of say, speak, talk, tell with possible
German counterparts. PSICL 6, 99-109.
Mel‟chuk, I. A. (1963). Machine translation and linguistics cited in James,
C. (1980). Contrastive Analysis. London: Longman.
Politzer, R. L. (1972). Linguistics and applied linguistics: Aims and
methods. Philadelphia: Center for Curriculum Development.
Sajavaara, K. (1985). Language processes in contrast: contrastive analysis
revisited. In
Stockwell, R. P., Bowen J. D. and Martin (1965). The grammatical structure
of English and Spanish. Chicago: University of Chicago Press.
Wardhaugh, R. (1970). The contrastive analysis hypothesis. TESOL
Quarterly 4 (2), 123-130.
Whitman, R. L. (1970). Contrastive analysis: Problems and procedures.
Language Learning 20 (2).
Wonderly, W. L. (1968). Bible Translations for popular use. London: United
bible societies
... Sorotan kajian Mohammad Hamad Al-khresheh (2016) menyokong pemilihanan teori kajian pengkaji kerana dapatan kajian beliau mengenai Teori AK adalah masih signifikan dan masih relevan sehingga ke hari ini. Begitu juga dengan sorotan kajian Mayada Zaki (2015), iaitu menjelaskan tentang kelebihan AK dalam perbandingan B1 dan B2. Selain itu, sorotan kajian ini turut memberikan Rajahan kepada pengkaji tentang cara menjalankan kajian perbandingan BA dengan BM melalui AK. ...
... Teori ini turut menjadi pilihan kajian lepas oleh Norizan Che Su (2017), Tajudin Nur (2016), dan Mariyati Haji Mohd Nor et al (2015). Sorotan kajian Mayada Zaki (2015) turut menyatakan sumbangan AK dalam P&P adalah keupayaan untuk meramalkan dan mendiagnosis kesilapan B2 pelajar serta reka bentuk ujian sebagai panduan kepada pelajar B1. Oleh yang demikian, sorotan kajian ini turut memberikan idea kepada pengkaji untuk memilih Teori Analisis Kontrastif Lado (1957) dalam menjalankan kajian perbandingan penulisan karangan BP -BM. ...
... Beliau juga menyokong bahawa dalam proses pembelajaran bahasa apabila terdapat masalah gangguan seperti "interlanguage", maka perkara yang mudah boleh menjadi susah dan masalah ini dapat diatasi dengan pendekatan AK. Bertitik tolak dari sorotan kajian inilah, pengkaji mendapat idea untuk menjalankan kajian AK ini dan seterusnya memilih pendekatan Mayada Zaki (2015) dalam menjalankan kajian perbandingan KP kata depan tempat BP-BM. Dapatan dari kajian Analisis Kesilapan digunakan untuk melihat kesalahan pemilihan KP dan seterusnya data tersebut digunakan untuk mengkaji persamaan dan perbezaan yang terdapat pada KP kata depan tempat "di" BP-BM. ...
... Sorotan kajian Mohammad Hamad Al-Khresheh (2016) menyokong pemilihanan teori kajian pengkaji kerana dapatan kajian beliau mengenai Teori AK adalah masih signifikan dan masih relevan sehingga ke hari ini. Begitu juga dengan sorotan kajian Mayada Zaki (2015), iaitu menjelaskan tentang kelebihan AK dalam perbandingan B1 dan B2. Selain itu, sorotan kajian ini turut memberikan gambaran kepada pengkaji tentang cara menjalankan kajian perbandingan BA dengan BM melalui AK. ...
... Beliau turut mencadangkan agar satu kajian dijalankan terhadap faktor-faktor kewujudan variasi bentuk kata sendi nama "di" dalam BP dengan pengkelasan yang teratur dan lebih jelas agar dapat membantu pelajar menguasai topik kata sendi nama BP dengan baik. Bertitik tolak dari sorotan kajian inilah, pengkaji mendapat idea untuk menjalankan kajian AK ini dan seterusnya memilih pendekatan Fatin Najlaa Malik et al (2019) dan Mayada Zaki (2015) dalam menjalankan kajian perbandingan KP kata depan tempat BP-BM fokus kepada KP kata depan tempat bagi negara dan bandar. Sehubungan dengan itu, dapatan dari kajian Analisis Kesilapan digunakan untuk melihat kesalahan pemilihan KP dan seterusnya data tersebut digunakan untuk mengkaji persamaan dan perbezaan yang terdapat pada KP kata depan tempat "di" BP-BM. ...
The claim that the best language-teaching materials are based on a contrast of the two competing linguistic systems has long been a popular one in language teaching. It exists in strong and weak versions, the strong one arising from evidence from the availability of some kind of metatheory of contrastive analysis and the weak from evidence from language interference. The strong version of the hypothesis is untenable and even the weak version creates difficulties for the linguist. Recent advances in linguistic theory have led some people to claim that the hypothesis is no longer useful in either the strong or the weak version. Such a claim is perhaps unwarranted, but a period of quiescence is probable for contrastive analysis itself.
This volume sets forth in a nontechnical manner the linguistic approach employed in writing instructional materials used in English-as-a-second-language programs at the English Language Institute of the University of Michigan during the 1940's and 1950's. Each section of this volume presents the principles or the assumptions underlying the choice, sequence, and handling of the materials of the "Intensive Course in English for Latin-American Students." Chapters include: (1) "On Learning a Foreign Language as an Adult," (2) "The Sounds: Understanding and Producing the 'Stream of Speech'," (3) "The Structure: Making Automatic the Use of the Devices of Arrangement and Form," (4) "The Words: Mastering Vocabulary Content," and (5) "Contextual Orientation." Appendixes contain"Step-by-Step Procedure in Marking Limited Intonation,""Lessons in Pronunciation, Structure, and Vocabulary from 'Ingles por Practica'," and "Outline of Materials of 'An Intensive Course in English for Latin Americans'." (RL)
In the present study I am adducing arguments against the assumption that there is a discrete level of grammar where all syntactic operation has stopped and the assignment of phonetic values to the then existing structure (the output of transformations) begins. The concept of ‘surface structure’ has received the following definitions in Chomsky's various works: ‘a proper bracketing of the linear, temporarily given sequence of elements, with the paired brackets labelled by category names (that is a labelled tree diagram, with such categories as Sentence, Noun Phrase, Verb Phrase, Noun, and a small number of others serving as labels)' (Chomsky, 1966: 21), ‘the organization into categories and phrases that is directly associated with the physical signal’ (Chomsky, 1968:25). These definitions take the physical event of speech as a starting point, and refer to the organization of the sentence on this level (cf. Chomsky, 1965a: 15). One finds, on the other hand, that in the Chomskyan system ‘surface structure’ is also defined in terms of derivation, as the last item in a finite sequence of phrase-markers, the output of the transformations (Chomsky, 1970: 53 f., 87), and such surface structures ‘are mapped into phonetic representations by the phonological rules’ (ibid.: 54) and are assigned intonational contours by rules of phonological interpretation (ibid.: 77). This is summed up in the formula ‘phonology: P n →phonetic representation’ (ibid.: 87, where P n is the surface structure, the last phrase-marker).