Connect, Grow, Thrive

ETAS Journal Editors' Choice Number 42

The mental lexicon and its implications for EFL teaching: Some insights

Stephan Kneubühler 

ETAS Journal Volume 37, Number 1 (Winter 2017), pp. 40-43

If you are looking for a short yet thorough article on the ‘mental lexicon’ and how this term took on new meaning for a young teacher-researcher when he met his new class for the first time, then look no further. Stefan Kneubühler’s article is the one. The commonplace experience of mixing up students’ names inspired the author to take a look at the processes involved in remembering words and to use the psycholinguistic perspective to get a better understanding of students’ lexical development. Is it easier to learn new words in one’s L1 or one’s L2? Stefan Kneubühler might just have the answer.

Trudy Krkoska

This article includes the text only. 

The mental lexicon and its implications for EFL teaching: Some insights 

Stephan Kneubühler 

On the first day of school after the summer holidays I was excited to finally welcome my new class. The books had arrived on time, my lesson plans were ready and I had even had time to practise my students’ names using the photo list their previous teacher had provided. There was really nothing to worry about when I greeted each student by shaking hands. “Good morning, Seraina. Morning, Jasmine. Good morning, Luca.” “I’m Lars.” “That happens to everybody”, actual Luca, waiting a little bit further in line for his handshake, commented with a smile. And it kept happening to me for the rest of the first week. This experience seems to be a good starting point to approach the question of how not only names but words in general are stored in the human brain. 

The mental lexicon and word association 

Mental lexicon is the psycholinguistic term for “a person’s mental store of words, their meaning and associations” (Richards & Schmidt, 2002, p. 327). In his attempt to understand and explain the mental lexicon, Carter (1998) points out that words do not exist in isolation. Their meanings are defined through sense relations they have with other words. McCarthy (1990) states that a total model of the mental lexicon “will have to be three-dimensional, with phonological nets crossing orthographic ones and criss-crossing semantic and encyclopaedic nets. This word-store is constantly being updated, new words are added, unused words may be forgotten, and new connections are being made” (p. 41). This is equally true for native speakers (NS) and L2 learners (McCarthy, 1990, p. 42). 

In her extensive work on the mental lexicon, Aitchison (2003) mentions word-association as one tool for the exploration of the human word store. There are different variations of word-association tests of which the underlying principle however remains the same: After a stimulus word is presented to the candidates they are asked to respond with the first word or words that come to mind. Depending on the form of the test, presentation and response can either be verbally or in written form. 

Through careful classification of the candidates’ responses, association patterns are thought to emerge. Traditionally, researchers have been concerned with three types of responses: paradigmatic, syntagmatic, and clang responses. 

Paradigmatic (also choice or verticalresponses are words that could perform the same grammatical function as the prompt word in a given sentence. Therefore, paradigmatic responses are usually from the same word class as the prompt word. This study considers three subcategories of paradigmatic answers: coordination, hyponomy, and hypernymy as well as synonymy. 

Words which are “on the same level of detail” (Aitchison, 2003, p. 86) are referred to as coordinates. The concept of coordination includes antonym as well as co-hyponomy and is often exemplified by words such as “cat” and “dog”. Hyponymy describes hierarchical relationships including more specific terms describing asymmetrical relationships of two or more words. This concept can be illustrated by the words such as “animal” and “cat”, where the hypernym is superordinate to the hyponym. In synonymy a distinction can be made between strict synonymy and loose synonymy. While strict synonymy refers to two words which can be used interchangeably in all contexts, loose synonymy describes relationships between words of similar meaning across many, but not necessarily all, contexts (Coulthard et al., 2000, p. 24). However, since stylistic differences limit substitutability there can be no such thing as, nor any need for, totally substitutable synonyms in an absolute sense. This is why the term synonym is here used to refer to both strict and loose synonyms. For example, 

In the basket she discovered a cat. (Synonym: kitty, moggy, feline

As shown above, synonyms are illustrative examples of how paradigmatic answers could perform the same grammatical function as the prompt word and as such are also referred to as vertical responses. 

Let us now turn to horizontal ways of word-association. Syntagmatic (also chain or horizontal) responses bear a sequential or collocational relationship to the prompt word and are usually, but not necessarily, from a different word class than the prompt word. Collocation refers to the phenomenon that some words tend to co-occur regularly and predictably with other particular words (Richards & Schmidt, 2002, p. 95). The term collocates here refers to the other words which typically co-occur. A useful example of a (lexical) collocation of the word ‘cat’ might be the word ‘hairball’.

The term multi-word item refers to phrases or groups of words which function as single lexical items and which native speakers usually decode as “chunks” (Coulthard et al., 2000, p. 62). To L2 learners multi-word units are often semantically opaque, that is, their meanings cannot be deduced from their constituent parts. However, as Carter (1998) points out, “the different degrees of possible fixity or ‘frozenness’, both syntactic and semantic should be noted” (p. 66). 

Word-associations solely evoked by phonological or orthographical similarities are labelled clang responses. Often they bear no overt semantic connection to the prompt word. Aitchison (2003) refers to the phenomenon that similar-sounding or similar-looking words are often confused with one another as the bathtub effect since it seems to be easier to remember the “head” and the “feet” of a word than the middle. Illustrated below are malapropisms, cases in which a similar sounding word has been erroneously selected and which can be understood as evidence that words are, to some extent, phonetically and orthographically organised in the mental lexicon. Below is a selection of malapropisms produced by Mrs. Malaprop in Richard Sheridan’s play The Rivals (1775). 

“Nay, no delusions to the past
- Lydia is convinced.” (= allusions

“You could have knocked me over with a fender.” (= feather)

“Your being Sir Anthony’s son, captain, would itself be a sufficient accommodation;”
( = recommendation

Finally, some responses are related to a candidate’s personal world knowledge, cultural background, interests, age or other personal factors to a greater degree than they are to semantic, collocational or formal features (McCarthy, 1990, p. 41). In this article they are referred to as encyclopaedic responses. McCarthy (1990) notes that especially bias words such as war evoke encyclopaedic responses. 

Research and research method 

In order to gain a better understanding of the mental lexicon and the lexical development of L2 learners, word-associations of lower and higher level L2 learners were compared with the responses given by native speakers (NS). The experimental procedure followed task 123 in McCarthy’s Vocabulary (1990, p. 152). 

The word-association activity consisted of two online surveys which were both administered in written form to all of the subjects. The first survey comprised eight pages. The introduction page explained the task and setting of the activity. 

The seven stimulus words book, become, it, microwave, strange, behind and get were presented on the following seven pages individually in order to avoid interference. There was no time limit for completing the survey, but since candidates were instructed to respond with the very first word associated, it took them no longer than five minutes to complete. In order to investigate the hypotheses mentioned in the previous section, stimulus words of both low and high frequency were chosen. For the exploration of a possible interrelation of learners L1 and L2 mental lexicon the false friend become was included in the list of prompt words. 

Categorising the responses of the initial pilot study raised questions about the reliability and validity of the word-association activity. In order to minimise the danger of subjective judgement during categorisation of the candidates’ responses a follow-up activity was added. Candidates had to explain why they had chosen a certain response in a post-test survey. The two steps of responding to the prompt words and explaining the response where separated deliberately in order to maintain the impulsive nature of word-association. 

A total of 56 subjects participated in the study: 32 Swiss L2 learners and 24 NS.
The participants were divided into three groups: 

Name, level (according to the Common European Framework of Languages) and number of subjects of the three different groups. 

Fitzpatrick (2006) points out the “observed differences in the association behaviour
of children and adults” (p. 321). Therefore it was appropriate to conduct the study with a group of NS comparable to the L2 learners in terms of age and stage of cognitive development. Thus, the results of the Swiss L2 learners were compared with the responses of a group of 24 Californian middle school students. 

This might imply that associations based on phonological or orthographical similarities play an important role at an initial stage of the language learning process but as learners become more proficient their importance shrinks and stagnates. 

As previously stated, the verb become was included in the study since it was expected to produce responses that would allow us to draw conclusions about the connection of L1 and L2 mental lexicons. This assumption however, was solely based on my own experience as a native speaker of German and language learner as well as a language teacher. Due to orthographical and phonological similarities the English verb become and the German verb bekommen for L2 learners often seem to be related. However, they differ semantically. Kœssler and Derocquigny (1928, cited in Gutknecht, 2003, p. 698) introduced the term faux amis, false friends in English, to refer to this phenomenon. Due to formal similarities, L2 learners appear to transfer their L1 knowledge about bekommen to the L2 verb become. As illustrated in Table 2, a considerable number of responses to the prompt word become indicate L1 interference since they would be rather expected to be elicited by the verb get than the verb become. 

All in all, it can be said that the responses of NS were much more similar to the L2 learners’ than would be expected based on past research. There are different explanations for this. One reason might be the choice of prompt words. As mentioned earlier, the word-association results of Aitchison (2003) are based on high frequency nouns and adjectives only, whereas the stimulus word in the present study are less homogenous and include pronouns, verbs, and prepositions as well as nouns and adjectives. 

Conclusion: Implications for teaching 

It is unarguably tempting to believe that word-association can reveal information about the development and organisation of the mental lexicon. However, as indicated earlier, previous studies have failed to produce consistent findings. What the vast amount of sometimes contradicting results nevertheless certainly shows is how complex and at the same time highly organised the mental lexicon seems to be. The mere quantity of individual r esponses to one and the same stimulus word demonstrates that word-association and therewith the organisation of the mental lexicon is a highly individual issue and should be therefore treated accordingly. 

Simply clarifying the meaning of a new vocabulary item by using the student’s L1 most certainly is not enough to fully incorporate them into the mental lexicon in a way that they can be retrieved later on. Translation might be one link but in order to meet the requirements of McCarthy’s (1990) dynamic, three-dimensional model with phonologic, orthographic, semantic, and encyclopaedic nets criss-crossing each other, students need to establish respective associations. 

The following vocabulary task was presented to the students as a while reading exercise for the book The Beach by Alex Garland. Different examples of L2 learners’ solution for the task illustrate how individually the task was tackled. By designing clusters for new vocabulary items students are encouraged to delve into their individual features and must address paradigmatic, syntagmatic, encyclopaedic, and clang associations for the new words. Whether activities like this promote the integration of new vocabulary items into a language learner’s mental lexicon is still to be investigated. 

In this article I began by describing how I kept mixing up two students as the point of departure for further exploration of the human word store. For the past six months I have now worked with my class and I have got to know each student. Since that first week of school I have not confused Lars and Luca again. Maybe words are a bit like people: very similar at first sight, very individual once you get to know them better. Words certainly seem to be much more than just their orthographical or phonological form, their translation, their meaning or use. The sum of all of these things and the connections between them appear to be what they are. But we have to spend time with them, get to know them, to be able to work with them. 


Aitchison, J. (2003). Words in the mind. Oxford, UK: Blackwell. Carter, R. (1998). Vocabulary: Applied linguistic perspectives. 

London, UK: Routledge. 

Coulthard, M., Knowles, M., Moon, R., Deignan, A. (2000). Lexis (2nd edition). Birmingham, UK: The Centre for English Language Studies, The University of Birmingham. 

Fitzpatrick, T. (2006). Habits and rabbits: Word-associations and the L2 lexicon. EUROSLA Yearbook, 6(1), 121-145. 

Gutknecht, Ch. (2003). Translation. In M. Aronoff & J. Rees-Miller (Eds.), The handbook of linguistics. Oxford, UK: Blackwell 

McCarthy, M. (1990). Vocabulary. Oxford, UK: Oxford University Press. 

NODE (1998). The New Oxford Dictionary of English. Oxford, UK: Oxford University Press. Retrieved January 1, 2017 from,https://en.oxforddic...

Richards, C., Schmidt, R. (2002). Dictionary of language teaching & applied linguistics (3rd edition). London,UK: Longman. 

Sheridan, R. B. (1775). The rivals: A comedy. Retrieved January 20, 2017 from