Crystal (1992) notes that “teachers need to be aware of a range of methods, in order to find the one most appropriate to the learner’s needs and circumstances…”, and it is “frequently necessary to introduce an eclectic approach, in which aspects of different methods are selected to meet the demands of particular teaching situations”. Some methods prioritise the teaching of “language structure (form)”, while others the teaching of “language use (function)”. The underlying FLL view behind the former form-centred teaching is Behaviourism, whereas, when language use is given priority, the Cognitive view underlies teaching. Accordingly, the “grammar-translation method”, the “direct method” and the “audio-lingual method” seem to lean towards the former, whilst the “communicative methods” and the “humanistic approaches” to the latter category.

According to Bibauw (2021), “traditional language teaching is not behaviourist” but “the contrary”. Bibauw (2021) identifies traditional language teaching with the grammar-translation method, and summarises its characteristics such as “using the mother tongue, explaining the grammar, and focusing on reading the literature”. Although Bibauw (2021) claims that “behaviourist language teaching does the exact opposite”, “using the mother tongue, explaining the grammar, and focusing on reading the literature” do not necessarily mean that the grammar-translation method does not include behaviouristic aspects. Crystal (1992) argues that in the Behaviourist view, “FLL is seen as a process of imitation and reinforcement”, which is also reflected in foreign language course books before the appearance of the direct method.

According to Bárdos (2000), teaching grammar can be deductive, inductive and comparative. The grammar-translation method approaches grammar in a deductive way, which means that the teacher explains the grammar rule first, then, the learners are provided examples, which is followed by practising the rule. Grammar practice involves transformation exercises, fill-in-the-gaps exercises and substitution exercises. Bárdos (2000) notes that the grammar-translation method did not care about the frequency and the usefulness of a word. Bárdos (2000) connects the dominance of the grammar-translation method with the teacher’s maximalism and having students learn 40-120 words from lesson to lesson. Bárdos (2000) refers to reading as the preferred skill of the method. Furthermore, he mentions breaking down grammar into its smallest possible constituents, which is characteristic of intensive reading, a favoured subtype of reading in the grammar-translation method. He regards the grammar-translation method as markedly text-centric, and the method focuses on reading original or translated texts and very rarely on writing. According to Bárdos (2000), the grammar-translation method favoured a learning style made up of analytic, closed and concrete steps.

Bárdos (2000) suggests that the history of modern foreign language teaching and learning could be seen as a way of getting rid of what the grammar-translation method entailed on language learners, which is the eternal use of bilingual dictionaries and the misbelief that every word of a language has a translation equivalent in the other language. Bárdos (2000) sums up the idea as a medieval phenomenon, which regarded cramming the dictionary as a token of defeating the seven-headed dragon. As Bárdos (2000) notes, the assumption that every word has an exact equivalent in the other is mostly characteristic of the grammar-translation method.

There have been several models about how vocabulary items are structured in a bilingual person’s memory. DeGroot (2013) emphasises the “unrealistic suggestion of the models of the bilingual mental lexicon” except Weinreich’s (1953, 1968) “coordinative model”. By “unrealistic suggestion” she means that the “L1 and L2 terms of a translation pair are implicitly assumed to have exactly the same meaning”. However, it is a widely accepted fact that “complete meaning equivalence of a pair of translations rarely exists”. “Shared meaning aspects” and language-specific aspects may coexist. DeGroot (2013) also notes that “different types of words … may differ in the amount of meaning they share with their translations”. Furthermore, she adds that “word meanings … change over time and differ between individuals”.

According to Fischer (2010), the role of translation in language education has shifted and traditional tasks like translation exercises have been phased out due to conflicts with communicative language teaching. She notes that while some teachers still use translation, the lack of guidance may lead to reliance on old methods. She also adds that there are arguments, however, that support integrating translation, emphasising its compatibility with communicative teaching principles, and, that translation enhances language skills and cultural knowledge. She concludes that empirical research is needed to assess its effectiveness compared to other tasks.

Bárdos (2000) explains the perfection of language teaching methods in the 20th century with the Hegelian thesis-antithesis-synthesis course of development and regards the communicative method as the prevailing thesis, whose antithesis has not appeared yet. If we set the prevalence of CLT aside and treated the grammar-translation method as any other, based on literature, we could summarise its main principles and compile a list of criteria the lessons of a language course book ought to follow.