Evolution of English Language Assessment: A Critical Account

1. Introduction

Educational curricula across the ‘secular territories’ of the world often claim to be promoting neutral and secular knowledge through textbooks and teaching in formal instructional settings. But looking into these documents and the practices critically may help us unveil the concept of ‘legitimate knowledge’; the political motivation for such construction; and the game of power, domination and marginalization they involve. Language learning being one of the most important parts of education has been exploited as an agency through which recursive and coercive practices like ‘selection’ and ‘judgment’ are carried out effectively. And thus, language policy of a nation is often found favouring the knowledge domain(s) of the power-holding section(s) of the nation. It does not democratically represent all sections of the society in terms of the languages, cultures, beliefs, political adherence, economic and social status, etc. This kind of practice may not be blatantly intentional. But one may wonder why and how such practices find their way into language policies sometimes in the name of ‘secularization’ or what Weber calls ‘rationalization’. This is actually a very clever way of constructing and moulding the consciousness of the dominated class(es) without being overtly resorting to mechanisms of social control. (Apple, 1979)

Historically proven to be a powerful social tool, language testing is used by language policy makers to promote a desired and favoured section of society as winners and accepted ones, and the rest as losers, failures and rejected ones. Though voices have been raised against the door-keeping nature of tests and attempts have been made to make testing student- and learning-friendly, we still continue to be acknowledged by our performance in high stake standardized tests like TOEFL and IELTS. However, a better picture of such undemocratic practices can be obtained if one looks into the history of English Language Assessment.  

2. Stages in Evolution

English language assessment has a history that runs parallel to that of English language teaching, and thus, it seems to have evolved in line with changes in approach to language teaching. First Spolsky (1975), and then Brown (1996), Bachman (2000), etc have tried to look at the developments in the field of language assessment through stages though all of them agreed that the divisions should not be interpreted as watertight compartments with exclusive features. Apart from using the major trends described by the above mentioned authors, I have also added a section on the recent advances in language assessment that includes the rise in awareness about social dimensions and exploration of alternative forms of assessment. 

2.1 The Beginning: English Language Testing and the British Stronghold

English language testing began in the 15^th century, the time when English Language Teaching was in its infancy. Henry V started an English Language policy according to which French was to be replaced by English as the language of royal correspondence. There arose the need of teaching English to people. But the decisions related to teaching and testing methods were taken by tutors. That did not lead to growth in teaching and testing since it was not allowed to spread among the mass. But after 16^th century when attempts were made to define and conceptualize language, serious attention was paid to developing methods of teaching English. With Johann Christian Fick’s ‘Practical English Course’ (1793) and John Miller’s ‘The Tutor’ (1797), ELT was on the track of productive research, theorization and experimentation. However, English language testing had to wait until 1913 to take the shape of modern standardized tests. The University of Cambridge sent papers to British colonies to examine 10 candidates in 1863 and continued to do so till the end of the century. The number of candidates kept growing and reached 1220 in 36 centres in different colonies by 1898. It is only in 1913, Certificate of Proficiency in English examination was started to give foreign-qualifiers a status of proficient user of the language. The hegemony of English and simultaneously, standardized tests has been continuing since then. During all these years, examinations focused on test items, and the manner of testing took a back-seat.

If the British started the linguistic imperialism through English Language Education, Americans were not far behind. Spolsky (1993) talks about how the US immigration department used English language test as a tool of control to rationalize the inflow of immigrant students to the country in the early 1920s. The ‘prognosis tests’ were seriously segregating and elitist in nature. The whole world does not seem to have broken free of the organized filtering practiced through language tests developed by the US and the UK. 

2.2 Grammar-Translation Approach and Testing

The pre-Lado period, i.e., the period before 1960s in testing bears a strong influence of Grammar-Translation Approach to language teaching. It is surprising that the newly emerging linguistic awareness in language pedagogy did not seem to have affected language testing in the first half of the twentieth century. The emphasis on detailed analysis of grammar rules and using this knowledge for translating sentences from L1 to English (as an L2 or a foreign language) and vice-versa confined language learning to memorization of rules and vocabulary. Consequently, language testing focused on the accurate and error-free reading, writing and translating ability of learners and also their ability to remember vocabulary in the target language. Testing oral proficiency and ability to use language for communicative purposes were never in the agenda of test-designers. But the tests were easy to design, and assigning marks was quite time-saving. A very good example of such tests could be The Charter's Diagnostic Language Test and the Pressey English Test which tested Grammar, Punctuation, Capitalization, and Sentence Structure. Lack of objectivity and statistical analysis marred the efficiency of these tests. Hence, this period is called ‘pre-scientific’ by Spolsky (1978).

If we look at the tests of this period a little more closely, we can realize that they indirectly favoured students from elite backgrounds. The overemphasis on the use and learning of classical texts; the amount of time needed to learn the structures; the affinity between the languages of the elites and the textbooks; the role of the teacher (mostly from upper classes) as the unquestionable authority in the classroom; etc are indications that students from socioculturally and economically disadvantaged background were never considered in sympathetic light by the policy makers. Moreover, these tests were used as means to declare students from such classes as mentally underdeveloped and thus, unfit for learning in formal settings.

2.3 The Structuralist Approach to Testing

The Grammar-Translation approach to testing was found inappropriate and ineffective and replaced by what Spolsky (1978) calls a ‘psychometric-structuralist’ trend in the 1960s. This trend bore the influence of behaviourists like Skinner and structural linguists like Fries and Bloomfield. Language learning was perceived as a process of habit formation, and language testing was a process of measuring language skills and elements of language at discrete levels. Once again, individual student and context were neglected in the name of science, objectivity, validity, reliability and precision.

During this period, students were encouraged to compete with each other as test scores got ultimate importance.  Individual abilities took the backseat. A fixed standard was created through discrete-point testing to test language ability of students. Passing and failing were prioritized over learning. All these led to what Foucault (1971) calls “appearance of a new modality of power in which each individual receives as his status his own individuality”, and in this framework individuals were confined to being ‘cases’ (Foucault, 1971). The students were the worst sufferers in this case because they were forced to confirm to a set of predicted behaviour. This, in turn, resulted in the suppression and loss of natural learning abilities of students.

2.4 Integrative Approach to Testing

The opposition to discrete-point testing gave rise to an integrative approach to language testing, i.e., a combination of psycholinguistic and sociolinguistic approaches. Influenced by cognitive psychology and learning theories, the psycholinguistic view of learning challenged the theories proposed by Saussure and Bloomfield. Chomsky’s theories about language and learning acted as a driving force. Language testing, under the influence of this theory, shifted its emphasis from linguistic accuracy to functional ability. Language tests adhered to problem-solving approaches and were expected to reveal what underlying rules the learners had internalized. The sociolinguistic views of language, on the other hand, were guided by Hymes (1972) who emphasized that the social context of a message is as important as its linguistic context. Accordingly, a language test was expected to measure one’s ability to use linguistic elements and skills communicatively and appropriately in a given social situation.

Even the inclusion of contextual tests could not solve the problems of language testing. The social and cultural dimensions of the tests were overlooked. The linguistic, cultural and social minorities were victimized by the contexts in which language items were tested. The only excuse for their alienation was the impossibility of accommodating a variety of contexts in any single test. This could be interpreted as a hidden agenda that promoted a ‘unilateral intellectual authoritarianism’. There is no need to mention who or which section of the society benefitted the most from it.    

2.5 Communicative Approach to Testing

Hymes’ model of “communicative competence” continued to guide the field of language testing in Europe and America till the 1980s. In 1980s and the early 1990s, the language testing models proposed by Morrow (1979), Canale and Swain (1980) and Bachman (1990) who emphasized on testing both, competence and performance of the learner gained in popularity and importance. Bachman (1990) defined language ability as a combination of two components: “language competence”, i.e., a variety of language knowledge and “strategic competence”, i.e., a set of metacognitive strategies. Bachman and Palmer (1996) took up from where Hymes had left. They argued that the construct and context of tests must be defined clearly; the materials and test tasks must be as authentic as possible; and real life situation must form the background of all test items. In addition, they asserted that a test must take into account and measure linguistic, sociolinguistic, discourse, and strategic competences.

Despite the inclusiveness and convincing nature of the communicative framework of testing, it has not become a part of the mass practice across the globe. This suggests that there is an obvious gap between theory and practice. The student-centredness which is often preached and unanimously agreed upon in any discussion related to language pedagogy has not really acquired practical propositions. We need to bridge this gap.

2.6 Critical Language Assessment and Alternatives in/ to Assessment

Language testing has been growingly recognized at as a political and ideological phenomenon deeply rooted in society and culture. Simultaneously, attempts have been made to explore these dimensions of language testing. The process of looking critically at testing started long time ago when Henry Latham (1877) criticized ‘encroaching power’ of examinations which he thought had a biasing effect on education. Almost a century later, similar kind of complaint was heard from Foacault (1977):

“…the examination is at the centre of the procedures that constitute the individual as effect and object of power, as effect and object of knowledge. It is the examination which, by combining hierarchical surveillance and normalizing judgement, assures the great disciplinary functions of distribution and classification, maximum extraction of forces and time, continuous genetic accumulation, optimum combination of aptitudes and, thereby, the fabrication of cellular, organic, genetic and combinatory individuality.”

Foucault might have been inspired by Paulo Freire’s (1970) masterpiece ‘Pedagogy of the Oppressed’. But both of them surely had some impact on the discipline of Applied Linguistics which was taking shape during the last part of 1970s. However, language education, especially language testing took a lot of time to accommodate critical perspectives in its discourse. The concepts of anxiety, bias, hegemony, democracy, marginalization, dominance, ideology, etc started to be taken seriously only in the 1980s, i.e., the time when Norman Fairclough (1989) published his monumental text ‘Language and Power’. This again shows that educational changes can be felt more strongly in the form of political visions than practical options.

During the 1990s, experts in language testing like Spolsky, Tim McNamara, Elana Shohamy, etc tried to go beyond the linguistic boundaries and look at language testing as a powerful educational tool that is used for social, political, cultural, and above all, ideological control. These advocates of fairness and ethics in language testing got solid support from the evolving discipline of Critical Applied Linguistics. The result is evident in concepts like critical language testing, democratic assessment (Shohamy, 2001), alternative assessments (Huerta-Macías, 1995), fairness (Kunnan, 2000), bias (Shepard, 1981), etc.

Looking critically at language testing has led to search for alternatives ‘in’ and ‘to’ assessment. Brown and Hudson (1998) list checklists, journals, logs, videotapes, audiotapes, self-evaluation, teacher observations, portfolios, conferences, diaries, self-assessments, and peer assessments as alternatives in assessments. These alternatives may be used as substitutes to what we call ‘testing’. They provide options that may help us move beyond technicalities of language testing and endear testing as a useful tool that can promote learning in formal educational centres. Moreover, they have paved way for a broad framework that can accommodate a variety of individual learning styles and preferences while acknowledging the identity and abilities of every single learner.  

3. Conclusion

This paper is an attempt at revisiting the history of English language testing in critical light. The aim is not to deny the necessity of tests. From the ancient example of the Shibboleth test in the Bible to the modern day tests like the Australian Dictation Test (1901), the Golden Rule Settlement (1940), Fruit Machine (1950), the Occupational English Test (1983), TOEFL, etc, language testing has proved to be more inhuman than human; more autocratic than democratic; and more a device of control than one promoting freedom. It is high time we mould it in such a way that it accommodates our multilingual and multicultural diversities; and individual preferences, personality traits, ideologies and beliefs.

References

Apple, M. W. (1979). Ideology and Curriculum. London: Routledge.

Bachman, L. F. (1990). Fundamental considerations in language testing. Oxford: Oxford University Press.

Bachman, L. F. (2000). Modern language testing at the turn of the century:  Assuring that what we count counts. Language Testing, 17, 1-42.

Bachman, L. F., & Palmer, A. S. (1996). Language testing in practice: Designing and developing useful language tests. Oxford: Oxford University Press.

Brown, J. D. (1996). Testing in language programs. Upper Saddle River, NJ: Prentice Hall Regents.

Brown, J. D., & Hudson, T. (1998). Alternatives in language assessment. TESOL Quarterly, 32(4), 653-675.

Canale, M., & Swain, M. (1980). Theoretical bases of communicative approaches to second language teaching and testing. Applied Linguistics, 1, 1-47.

Fairclough, N. (1989). Language and Power. London: Longman.

Foucault, M. (1977). Discipline and Punish: The Birth of the Prison. London: Penguin Books.

Freire, P. (1970). Pedagogy of the oppressed. New York, NY: Herder & Herder.

Huerta-Macias, A. (1995). Alternative assessment: Responses to commonly asked questions. TESOL Journal 5, 8–11

Hymes, D. (1972). On Communicative Competence. In J.B. Pride & J.Holmes (Eds.). Sociolinguistics. Harmondsworth, England:Penguin Books.

Kunnan, A. J. (Ed.). (2000). Fairness and validation in language assessment:  Selected papers from the 19th language testing research colloquium, Orlando. New York: Cambridge University Press.

Latham, H. (1877). On the Action of Examinations Considered as a Means of Selection Cambridge: Deighton, Bell.

Morrow, K. (1979). Communicative language testing: revolution or evolution? In Brumfit, C. J. and Johnson, K., (Eds.), The communicative approach to language teaching. Oxford: Oxford University Press, 143–57.

Shepard, L.A. (1981). Identifying bias in test items. In B.F. Green (Ed.), New directions in testing and measurement: Issues in testing-Coaching, disclosure, and test bias. San Francisco: Jossey-Bass, 79-104

Shohamy, E. (2001). Democratic assessment as an alternative. Language Testing 18, 4, 373 –392.

Spolsky, B. (1978). Introduction: Linguistics and language testers. In B. Spolsky (Ed.), Advances in language testing series: 2. Arlington, VA: Center for Applied Linguistics

Spolsky, B. (1978). (Ed.). Approached to Language Testing. Papers in Applied Linguistics. ERIC ED 16548

Spolsky, B. (1993). Testing of English of Foreign Students in 1930. ERIC 2002.

Spolsky, B. (1975). Language testing – the problem of validation. In L. Palmer & B. Spolsky (Eds.). Papers on Language Testing 1967-1974. Washington, D.FC.: TESOL. 147-53.