Vocabulary Learning Strategies and Foreign Language Acquisition

Multilingual Matters

Contents

Factors Affecting Vocabulary Learning and Acquisition

Despite the abundance of research on vocabulary acquisition that has been conducted by linguists, psychologists and theorists of L2 acquisition, there is still no generally accepted theory of vocabulary acquisition (for further discussion, see Meara, 1997). This fact may be partially attributed to the lack of cooperation or agreement among experts. On the one hand, psycholinguists have a particular interest in vocabulary development and exploration of the formal models of vocabulary acquisition, and ignore the L2 vocabulary literature because it is model-free. Applied linguists, on the other hand, are mainly concerned with the descriptive aspects of vocabulary and do not draw on existing psycholinguistic models of bilingual lexicon even when this implies an immediate pedagogical significance. Differences in the research focus have caused the two fields to develop at different rates, which has led to an even larger gap between them. It is, therefore, extremely difficult to list all the significant factors and the ways in which they influence vocabulary acquisition. In this section, a selection of the factors most frequently discussed in the relevant literature is presented.

Linguistic Features of Lexical Items

When it comes to linguistic features of lexical items, several issues need to be taken into consideration. To begin with, there is the problem of defining a 'word'. Intuitively, vocabulary could be defined as a 'dictionary' or a set of words. This general view is reflected in the lexicographical approach to the traditional way of listing words in a dictionary. However, it is obvious that for linguistics and L2 acquisition theory this interpretation is far too simplistic and limited. Linguists' attempts to specify what speakers of a language traditionally regard as a 'word' have resulted in so many formally different definitions of this term that their number alone suggests the complexity of the problem.

Firstly, according to the orthographic definition, a 'word' is '... any sequence of letters (and a limited number of other characteristics such as hyphen and apostrophe) bounded on either side by a space orpunctuation mark' (Carter, 1992: 4). Its flaw is not only its limitation to the written language, but the fact that it is formalistic, inconsistent and incomplete because it neglects differences in meaning and the issues of polysemy, homonymy, grammar functions, etc.

Secondly, based on semantics, a word can be defined as the smallest meaningful unit of language (Carter, 1992). As there is still no satisfactory definition of what 'meaning' is, i.e. what is the relationship between the linguistic sign and what it denotes outside the language, this definition is not reliable enough. Namely, some units of meaning consist of several words (e.g. bus conductor), for some the meaning cannot be determined without looking into their function in structuring and organising information (e.g. if, but), and certain 'integral' parts of words cannot stand on their own even if we know their meaning (e.g. the prefix 're-' in retell).

Thirdly, by the same token, the definition that restricts a word to a single stressed syllable allows for many exceptions: words like if and but do not have a stress, and bus conductor would be regarded as a single word in this view.

Next, Bloomfield's definition, according to which a word is a minimal free form, i.e. the smallest form that has a meaning when standing on its own (Škiljan, 1994), encompasses most of the categories and, without excluding further reduction of forms, provides a word with a degree of stability. Again, the problem of marginal cases arises and undermines every attempt to define a word in a formalistic way: firstly, items like a and the appear only in contextual relations to other words and secondly, idiomatic expressions, which consist of several orthographic words and cannot be reduced without radically changing their meaning (Carter, 1992).

Furthermore, McCarthy (1994) claims that a word, as a free meaningful unit of language, must contain at least one potentially freestanding morpheme. From this view a conditional definition of a word may be derived: a word is a combination of morphemes that comprise a firm unit suitable for the formation of higher level units (Škiljan, 1994). In addition, in Carter's view (1992), one of the greatest problems of defining a word, along with the abovementioned constraints, is the fact that words have different forms that would not intuitively be regarded as different words. Moreover, words can have the same form with completely different and unconnected meanings.

Finally, by way of attempting to solve this problem, a neutral term lexeme or lexical unit has been introduced. It is an abstract unit that includes various orthographic, phonological, grammatical and semantic features of a 'word'. Thus, this term covers inflections, polysemy, as well as multi-word items with different degrees of fixedness, such as compounds, phrasal verbs, and idioms. The difference between holistic multi-word items and other kinds of strings (i.e. multi-word inflectional forms, such as verb phrases are going or has been chosen) may be determined by applying the following criteria: institutionalisation or lexicalisation (the degree to which a multi-word item is considered as being a unit by the language community), fixedness (the degree to which a multi-word item is frozen as a sequence of words) and non- compositionality (the degree to which a multiword item cannot be interpreted on a word-by-word basis, but has a specialised unitary meaning) (cf. Moon, 1997: 44).

The second issue that needs to be discussed arises from the lack of an unambiguous and universally accepted definition of a word: vocabulary of any language consists of a wide range of lexical forms. Thus, many linguists and theorists of L2 acquisition agree that vocabulary is made up of a variety of forms, such as morphemes, both free and bound (e.g. laugh, or the prefix un-), their combinations, i.e. derivatives (e.g. laughter, unbelievable), compounds (e.g. bus conductor), idioms, i.e. units that cannot be reduced or changed, and whose meaning cannot be retrieved from individual meanings of their components (e.g. to bite the dust), and other fixed expressions, such as binomials and trinomials (e.g. sick and tired; ready, willing and able), catchphrases (e.g. they don't make them like that any more), prefabricated routines or prefabs (e.g. if I were you), greetings (e.g. How do you do?) and proverbs (e.g. It never rains but it pours). This list of formal categories indicates a tremendous heterogeneity and a wide range of lexical items, but is by no means complete and absolute, nor are the categories strictly demarcated: their overlap is inevitable. It is this aspect that places vocabulary on the boundaries between morphology, syntax and semantics.

The third issue takes into consideration the fact that lexical items can hardly be viewed in isolation from each other, for they enter, semantically speaking, into various relations. These include hyponyms (lexical items within the same semantic field, i.e. at content level), synonyms (two or more lexical items that have the same or nearly the same meaning but different form), antonyms (lexical items of opposite meanings) and homophones (lexical items that have the same form but different meanings).

Meaning can be studied by means of the so-called componential analysis, which is based on the assumption that the meaning of a lexical item can be broken down into a set of meaning components or semantic features. The meaning of a lexeme is determined by a number of distinctive semantic features, namely their absence (marked by '–'), presence (marked by '+') or irrelevance for the definition of a lexeme's meaning (marked by '[+ or -]'). This approach shows which features of lexical items from the same semantic field overlap or differ, and is therefore suitable for the exploration of synonymy. A disadvantage of componential analysis is not only its failure to cover all meanings, but also the fact that it reduces the meaning components to binary oppositions that cannot always be precisely determined, and the fact that it may result in an indefinite list of a lexical item's relevant features.

The above-mentioned cases exemplify a paradigmatic relationship. This is the relationship between a lexeme and other lexemes that could be substituted for it in a sentence. A different type of relationship which lexemes enter into – called a syntagmatic relationship – is characterised by linear sequencing of lexemes. Such combinations of lexemes, however, are restricted. These restrictions (or 'collocations') determine which lexical units may be selected to form semantically acceptable combinations of two or more syntactically combined lexical units. Some collocations are entirely predictable (e.g. blond and hair); some lexical items have a wide range of collocations (e.g. letter collocates with alphabet, box, post, write, etc.), and some lexemes appear in so many different contexts that it is practically impossible to predict all of their collocations (e.g. verbs like have or get). To be noted is the fact that collocations differ from free associations of ideas: associations are highly individual, whereas collocations are lexical connections established in the same way by all speakers of a language. The study of collocations can be effective if it is conducted on large amounts of data, which is inevitably associated with corpus studies, because collocations are not merely random combinations of lexical items, but are part of their meaning in the broadest sense of the word (Moon, 1997).

Finally, other factors influence the learning of a lexical item and make the acquisition of vocabulary difficult. According to Laufer (1997), the factors that affect the learnability of lexical items include pronounce- ability (phonological or suprasegmental features), orthography, length, morphology, including both inflectional and derivational complexity that increase the vocabulary learning load, similarity of lexical forms (e.g. synforms, homonyms), grammar, i.e. part of speech, and semantic features (e.g. abstractness, specificity and register restriction, idiomaticity and multiple meaning). Table 1.1 gives an overview of the intralexical factors and their effect on vocabulary learning (facilitating factors, difficulty-inducing factors and factors with no clear effect) (Laufer, 1997: 154).

The Influence of First and Other Languages

L2 vocabulary acquisition is different from L1 vocabulary acquisition because an L2 learner has already developed conceptual and semantic systems linked to the L1. This is why L2 acquisition, at least in its initial stages, often involves a mapping of the new lexical form onto an already existing conceptual meaning or translational equivalent in L1. The role of L1 in this process varies depending on the degree of equivalency between languages: although in some cases it may facilitate the acquisition or use of L2 lexical items, in others it will create an obstacle. This may occur in the process of acquisition, in recalling and using previously learnt lexical items, or in attempts of constructing a complex lexical item that has not been learnt as a unit. By making cross-linguistic comparisons (i.e. by contrastive analysis) one can often predict difficulties caused by interference of the L1 that learners may encounter when learning the target language. Namely, the learner's approach to L2 learning is based on an 'equivalence hypothesis': 'the learner tends to assume that the system of L2 is more or less the same as in his L1 until he has discovered that it is not' (Ringbom, 1987: 135). The learner's readiness to transfer may also be influenced by his perceptions of linguistic and cultural distance. Forming a kind of equivalence hypothesis enables learners to learn an L2 without having to go all the way back to learning how to categorise the world. However, equivalence hypothesis may fail and lead to erroneous conclusions because of the following reasons (Swan, 1997):

• lexical units in two languages are not exact equivalents (i.e. there is more than one translation);

• equivalent lexical units in related languages have different permissible grammatical contexts;

• equivalents belong to different word classes;

• equivalents are false friends;

• there are no equivalents at all.

Coping with these problems may be overwhelming, and the learners tend to avoid such 'difficult' lexical items, especially if there is a semantic void in the L1. A possible explanation is that in such cases there is no foundation on which L2 knowledge may be built (Gass, 1989).

Finally, the L2 learner, unlike the child acquiring its L1, cannot significantly expand his or her vocabulary solely through exposure to the language input. The exposure to L2 input is often limited to the classroom context. The input may be increased by reading (cf. Ellis, 1997) or listening (Rivers, 1981) in the target language. But these activities, although undoubtedly useful, do not guarantee the development of rich vocabulary. Similarly, formal teaching of vocabulary has its limitations, for, as Rivers (1981: 463) claims, 'vocabulary cannot be taught'.

The Incremental Nature of Vocabulary Acquisition

Knowledge of an L2 lexical item consists of several components. Generally, it is characterised by several dimensions of word knowledge (i.e. phonological and orthographic, morphological, syntactic and semantic) and by knowledge of conceptual foundations that determine the position of the lexical item in our conceptual system. Finally, it inevitably includes the ability of productive use, i.e. efficient retrieval of the lexical item for active use.

Ideally, knowledge of a lexical item would include all of the abovementioned dimensions and would be reflected in the ability to react in the manner of an educated native speaker. However, knowledge of a lexical item is not an 'all-or-nothing' proposition; it is rather to be conceived of as a continuum of knowledge at whose ends, according to some theoreticians, the receptive and productive knowledge is placed. It can be concluded that even partial knowledge represents a degree of knowing a lexical item. The initial degree is elementary knowledge, such as the visual recognition of a lexical item in a context that still does not enable a learner to produce it. Higher degrees of knowledge, close to productive knowledge, would suggest, for example, knowledge of multiple meanings of a polysemous lexical item or its collocations, etc. Whereas interpretation requires only as much information as is necessary to distinguish a lexical item from all other possibilities, production requires more information, which may even include the aid of an adequate stimulus (e.g. context) (Melka, 1997).

The Role of Memory in Vocabulary Learning and Acquisition

The role of memory is crucial in any kind of learning and vocabulary learning is no exception. According to the above-described continuum, learning of lexical items is not linear. Learners, without fail, forget some components of knowledge. In both long-term and short-term memory forgetting takes place in a similar way. When obtaining new information, most of it is forgotten immediately, after which the process of forgetting slows down. On the basis of available research results, Thornbury (2002) has compiled a list of principles that facilitate the transfer of the learning material into the long-term memory. These include multiple encounters with a lexical item, preferably at spaced intervals, retrieval and use of lexical items, cognitive depth (cf. Schneider et al., 2002), affective depth, personalisation, imaging, use of mnemonics and conscious attention that is necessary to remember a lexical item. A proper understanding of the role memory plays in vocabulary acquisition has an immediate practical value: as lexical knowledge is more prone to attrition than other linguistic aspects (Schmitt, 2000), the learning and teaching of vocabulary needs to be planned following the above mentioned principles if it is to be efficient.

The Organisation and Development of the Second Language Mental Lexicon

L2 vocabulary development is also influenced by the organisation of the mental lexicon. The mental lexicon is 'a memory system in which a vast number of words, accumulated in the course of time, has been stored' (Hulstijn, 2000: 210). This system is seen to be organised and structured, because it is the only possible explanation for the fact that people can, at an astonishing rate, in a vast quantity of lexical items stored in the memory, recognise and retrieve the lexical item they need to express what they want. Human memory is very flexible and it can 'process' a large quantity of data, but only if it is systematically organised.