English compounding: a pedagogical approach

Written by Deak Kirkham

Download a PDF version of this article

1. Introduction

In Andrew Niccol’s poignant 2005 film, Lord of War (Lord of War 2005), Nicolas Cage’s cool, calm and condescending arms dealer, Yuri Orlov, tussles with Eamonn Walker’s fictional dictator of Liberia, André Baptiste. Ostensibly their disagreements are about the supply of lethal weapons from Orlov to Baptiste; however they also disagree about word formation. Regarding this latter, on two occasions, Orlov comments on Baptiste’s choice of language form in (mis?)forming a compound noun.

Dialogues 1 and 2

It’s a clever motif. Perhaps it expresses Orlov’s detached condescension; alternatively (or simultaneously), it portrays Baptiste as a man without care for the rules. It may also act as a metaphor for the emotional disconnect both men have for the wars they resource and enact. However, it also demonstrates something important about human language(s): existing words can be combined in various ways to create alternative meanings. Of course, human languages draw on a range of resources to express meanings: in rare cases they go lexical and invent a new word (‘twerk’, ‘quark’); sometimes the syntax emerges a new structure or an existing structure extends (‘I’m X-ing it’ is currently highly productive: ‘I’m skydiving it this weekend; ‘I’m flapjacking it all Sunday’); metaphor is a third means of meaning creation involving a new use of a familiar word (‘sex up the dossier’); on other occasions they borrow from another language (‘a wiki’); names of people and places may also lend themselves to new meanings (‘Kafkaesque’, ‘Pythonesque’, ‘sadistic’, ‘coach’). But very often, as in the above dialogues, they just use their current linguist resources and create new words from existing words (‘miniskirt’, ‘internet’, ‘wireless’, ‘cupboard’, ‘morphophonological’ and ‘idiolect’); these are compounds.

Within linguistics, with respect to the ‘warlord’ / ‘lord of war’ alternation, the latter structure is phrasal: words combine to form a phrase. The former structure however is a compound which is defined here, following Bauer (2003: 40) to be the formal process of combining two or more lexemes such that the result is also a lexeme. In this sense, compounding is a particular type of word formation (i.e. it results in a ‘word’) and differs from both syntax (broadly conceived) and from other word formation processes (as well as from other processes like metaphor) as per Figure 1 [1].

Figure 1: Methods of forming new meanings in human languages

Figure 1: Methods of forming new meanings in human languages [2]

Before turning to particular types of compounding in English, we first unpack a little more the claimed distinction between compounding (combining words to make a word) and syntax (combining words to make a phrase). To linger with the opening example, the compound and phrase seem to mean very similar things: ‘warlord’ and ‘lord of war’. ‘War’ is non referential [3] (c.f. ‘Crimean Warlord’ which can only mean ‘a warlord (non-capitalised) from Crimea’, not ‘a lord of the Crimean War’ (capitalised); and ‘lord’ denotes ‘powerful person’ not ‘aristocrat’. However, a number of key differences between compounding and syntax can be identified (adapted from Spencer 1991). One key syntactic differentiator is the inability of the two elements of ‘warlord’ to be separated by any intervening element (Lieber & Štekauer 2009; Ryder 1994). In ‘lord of war’ (etc.), modifiers can be added to either key element as follows:

(1)       Lord of war                à        Imperial lord of war

(2)       Cup of coffee                        à        Big cup of delicious coffee

As Ryder (1994: 14) notes, no such intervention by any element can occur in warlord: *‘war imperial lord’; *‘war English lord’ and *’coffee big cup’. However, ‘imperial warlord’ and ‘English warlord’ are both acceptable. Connected to this is the inability of the compound to pluralise the first nominal element (or indeed take inflection (Lieber & Štekauer 2009)): *‘warslord’ is not English whereas ‘lord of wars’ is [4].

A second differentiator is stress (Spencer 1991). In the compound, primary stress occurs on the first element of the compound (‘wár lord’) whereas in adjective-noun combinations ‘English lord’, the stress falls on ‘lord’. This stress phenomenon is seen in the following alternations.

Figure 2: examples of stress pattern variation in phrasal versus compound word combinations

Figure 2: examples of stress pattern variation in phrasal versus compound word combinations

A further example of how stress plays a role in compounding can be seen in multi-word compounds. Take Spencer’s (1991) example of ‘government pay review committee’. This is ambiguous between the following two readings:

(3)       a [review committee] looking at [government pay]

(4)       a committee tasked with undertaking a [pay review] either of government or assigned that same task by government

In (3), ‘review’ receives stress, thus marking it as the first element of a two element compound; in (4) ‘pay’ receives the same stress. Moreover, in the reading in (3), no element may intervene between ‘review’ and ‘committee’ on the one hand and ‘government’ and ‘pay’ on the other c.f. *a government annual pay review public committee  versus a annual government pay public review committee.

Finally, at a semantic level, ‘warlord’ and ‘lord of war’ do not mean the ‘same thing’ and are not used in the same way. This, perhaps, is the (linguistic) essence of Orlov’s oral corrective feedback to Baptiste: effectively it’s a case of ‘we just don’t say it like that’ (or in Widdowson’s (1978) terms, a violation of usage). More specifically, this raises the issue of the (at least [5]) three ways in which nouns can be combined in English: NN, N’s N and N preposition N. These are exemplified in Figure 3 with a partial commentary on the ‘N part of N’ rule adapted from Swan (1995).

Figure 3: Three ways of combining nouns with example and commentary partially from Swan (1995)

Figure 3: Three ways of combining nouns with example and commentary partially from Swan (1995)

As the subsequent detailed discussion in Swan (1995) makes clear, the ‘part of’ rule is only one of several rules distinguishing these three noun formation processes, and there are, in any case, exceptions to it (‘roof of a house’ not ‘houseroof’ [6]). The point is this: whatever complexities NN compounding may contain, the pattern itself is complicated further by being one of three possible patterns in total.

With the phenomenon of compounding at least ostensibly defined, this article will now proceed to discuss the abundance and complexity of English compounding and place this in the context of English language teaching. Given the complexity that will be mapped out, it is argued that the learner is faced with a high degree of challenge. This challenge is not made easier by a somewhat unordered and non-comprehensive approach to compounding (and indeed other morphological phenomena) in the language leaning textbooks and support materials. In a review of standard textbooks (see Appendix), the author found little or no mention, certainly discussion, of compounding. Moreover, while finding more space for compounding, standard support materials do not offer much in the way of exercises, a graded syllabus or, again, a full linguistic explanation. The current article will therefore attempt to address this gap for language teaching professionals by presenting certain compounding phenomena in English in a systematic way, and then discussing the pedagogical approaches to them. In this way, the article hopes both to shed light on compounding as a linguistic phenomenon, and offer possible approaches to it in the contemporary language classroom. Throughout the discussion the question of the application of linguistics to language teaching will be had. Specifically, certain linguistic notions (headedness in compounds, semantic relationships and metaphor) will be given consideration. The aims of the article, then, are threefold:

  1. To unpack certain aspects of the English compounding system and to demonstrate their regularity and systematicity;
  2. To consider what aspects of the linguistics (broadly conceived) of these processes may be relevant to (certain levels of) the language classroom thus contributing to the discussion on the place of linguistics in language teaching;
  3. To suggest what pedagogical approaches may help learners to approach these structures thus addressing the gap in teaching materials and perhaps teaching practice.

2. Noun-noun compounds

2.1. The phenomenon

Noun-noun compounds (NNCs) are ubiquitous in English: computer screen, teacup, exercise machine, eyeglasses, screensaver, textbook and shirt sleeve. Their abundance is such that they can be grouped with some degree of systematicity. Figure 4a presents a large grouping based on the second element ‘book’; figure 4b exemplifies smaller sets with a shared second element.

Figure 4a: A partial set of noun-noun compounds with ‘book’ as the second element

Figure 4a: A partial set of noun-noun compounds with ‘book’ as the second element

Figure 4b: sets of compounds with the same right-most element

Figure 4b: sets of compounds with the same right-most element [7] [8]

We begin with two observations regarding similarity. Firstly, these sets of NNCs are identical in terms of their formal make-up (they consist of only two nouns, both in singular form [9]). Aside from the relatively trivial issue of the tripartite variation in written form between NN, N N and N-N, the syntax is simple and straightforward. Secondly, semantically, the second (right-most) element is always the semantic head of the compound: a poetry book is a kind of book; a throne room is a kind of room. This type of compound is called an endocentric compound: the semantic head is inside (endo-) the compound [10].

Against these similarities a more intriguing semantic difference: the kinds of meaning relations which obtain between the two elements differ widely, even within the sets. A storybook, for example is a book which contains (usually only one) story. A workbook, however, does not contain work (whatever that means); it is a book for working in, or in which work can be done. Again, a poetry book is a book containing almost certainly many (perhaps unrelated) poems whereas a science book is a book about science. A guidebook, finally, is a book which functions as a guide (or is it a book for a guide?). Clearly, also the second element, ‘shop’, is sometimes (variously) metaphorical (‘sweat shop’; ‘cop shop’). The metaphorical and perhaps culturally-bound meaning of certain compounds presents a further level of complexity to the phenomenon.

The issue of semantic unpredictability (and that of metaphor) appear also in compounds with shared first elements. Consider Ryder’s (1994) three NNC examples with ‘gold’ as the initial element: goldfish, gold-digger; goldsmith. In the first, the fish ‘resembles’ or has the quality of gold; in the second, gold is a metaphor for wealth and the meaning connection is around seeking wealth perhaps in a dubious way; in ‘goldsmith’ ‘gold’ is a material worked with. Noun-noun compounds, then, have very similar formal surface properties (N + N), but quite different, and unpredictable, semantic relationships exist between those nouns. Figure 5 offers a more detailed example in which some of these semantic relationships are now unpacked.

Figure 5: Some noun-noun compounds with ‘ship’ as the second element

Figure 5: Some noun-noun compounds with ‘ship’ as the second element

Note that despite the table’s attempt to group these ship compounds into sets, at least one remains ambiguous (‘slave ship’). More importantly, the list of semantic relationship types on the right is not complete. Further types are now offered (adapted from Girju et al. 2005 and Nastase et al. 2006) with noun-noun examples of each [11].

Figure 6: types of semantic relationships adapted from Girju et al. (2005)

Figure 6: types of semantic relationships adapted from Girju et al. (2005)

All aspects of NNCs discussed so far (the ostensibly simple formal order; the right-hand semantic headedness and the variable semantic relationships) can represent a learning barrier to speakers of certain languages. In Arabic, for example, the nearest equivalent to English NNCs is the idafah construction (Brustad et al. 2004). Although often cited as a type of NNC (Brustad et al. ibid., p.31), it differs along all three parameters. Firstly, headedness of the idafah is the inversion of the English order as in (5):

(5)       baytu al-        talabati

    house DEF-   student.PL

    the students’ house or a the student hostel

The Arabic compound in (5) is endocentric in the sense given above but the head is the initial element. This accounts at least in principle for Arab L1 speakers producing such compounds as ‘race car’ for ‘car race’. Secondly, as (5) also indicates, the idafah construction does not have to be a concatenation of two bare nouns; the second noun (only) can take the definite article. This is a difference from English where no element in an NNC can intervene between N1 and N2. Finally, although there can be some semantic ambiguity in the idafa (again, in (5), two English renderings are given), the relationship is always one of possession of some kind, not part-whole or make-produce. In this sense, the idafah is more semantically predictable than English NNCs.

As Spencer (1991: 312) points out, French L1 speakers have different but no less opaque challenges when approaching English NNCs. French has no formal equivalent, the only noun combination type available being syntactic and of the pattern noun + preposition + noun as in (6)

6a        un chemin-de-fer

     a road of iron


6b        le mise-au-point

     putting in focus [12]


These, as with Arabic have the opposite headedness to English. Another type of compound is common in French but again it not an NNC:

7a         porte-parole


      ‘A spokesperson’

7b         porte-monnaie


     ‘A purse’

7c         pince-nez



Thus, French has compounds, and can compound nouns, but not in the same way as English NNCs. The above is a brief introduction to the form and semantics of noun-noun compounds in which some aspects of the language-internal complexity has been presented along with some examples of L1 to L2 transfer barriers. We now turn to the pedagogical applications of this type of English compounding.

2.2. Pedagogical implications and applications

As noted above, one issue for English language learners of certain L1s is the ostensibly simple issue of headedness. This might be approached through a systematic noticing exercise (Schmidt 1990), observing of sets of compounds with shared first (goldfish, gold-digger) or second elements (swapshop, talking shop). With level-appropriate lexical items, such an exercise might be used fairly early in a syllabus to introduce the notion of an NNC and the notion of headedness. The same exercise might be repeated at a higher level to introduce the notion of metaphor through NNCs. A possible schematic worksheet is provided (examples partially from Ryder 1994) with low frequency lexemes in which metaphorical examples are square bracketed. A simple picture matching exercise of whatever form (pelmanism; placing pictures around the room) might introduce the terms which might then might easily be worked into a text to allow contextualisation. Learners would then be able meaningfully to guess what the compound might mean, or if not, realise that it is likely metaphorical.

Figure 7: Shared first element NNCs versus shared second element NNCs

Figure 7: Shared first element NNCs versus shared second element NNCs

A second more complex issue concerns the semantic opacity of noun-noun compounds as part of the learning challenge for second language users which derives from the lack of any grammatical clue as to what kind of relationship the two nouns have to each other. Here, there is an argument for the relevance, perhaps necessity, of the teaching of the relevant linguistic concepts in the language classroom. How might this be done? One option is a matching exercise perhaps using with the ‘ship’ example, above and a list of possible semantic relationships as in Figure 6. The role of the teacher will be to clarify and offer further examples of the semantic relationships. This can be done by offering fuller phrasal examples of any semantic terminology which causes difficulty. It is debateable whether the standard terminology need be used in labelling the types of semantic relationship (e.g. part-whole and attribute-holder). It may suffice to provide informal names for the relationships.

A more linguistically rich but metalinguistically simple way of approaching this puzzle is offering phrasal paraphrases of the various semantic categories. Thus, the causation relationship (‘flu virus’, ‘exam anxiety’) might be glossed with the phrase ‘N1 gives rise to / causes N2’. Similarly, the agentive structures can be glossed with ‘by’: ‘investigation by the government’; ‘protests by students’. Of course, phrases such as ‘give rise to’ will need to be taught but once taught, this can be employed in some kind of noticing activity in order to hone judgements and develop intuitions.

For higher level students, another awareness-raising approach might be to capitalise on the ‘government pay review committee’ type of string which allows two readings, disambiguated from each other via stress. This involves the notions of ambiguity and stress as well as the semantic relations discussed above. Teachers might introduce such examples (some of which are given below) and invite students to attempt to ascertain tow readings and assign stress to them. Students might also be asked which is more likely as a reading, and why, and then finally (as a productive activity) write a short paragraph containing the compound in one of its two readings, disambiguated by that co-text.

Figure 8: Further examples of NNNN strings with two interpretations

Figure 8: Further examples of NNNN strings with two interpretations

A final issue in which awareness raising might be relevant is the metaphorical nature of many NNCs. This has been touched on already. As is well known, metaphors vary along a continuum of opacity, with cultural and diachronic interference to a great or lesser extent. Other linguistic features can also play a role such as rhyme, as in ‘copshop’ and ‘swapshop’. In NNCs, either or both elements may be metaphorical (‘sweatshop’). In principle, of course, either the metaphorical or the compounding aspect of metaphorical NNCs could be emphasised in the teaching roll-out.

The above exercises are primarily awareness-raising and noticing activities with the aim of familiarising learners with the relevant linguistic concepts; they are receptive activities which aim to illustrate the concepts underpinning NNCs. We move now to productive activities, focussing on those which have an element of language creativity and language play, activities which may be categorised as ludic, a concept discussed in both education literature (Cekaite & Aronsson 2005[13]; de Castell 2011; Huizinga 1950; Kirkham 2015; Kolb & Kolb 2009) and psychological literature. Ludic activities have been defined by Huizinga (op. cit.) as those activities which are: a) free; b) take the player out of real life; and c) are bounded in time and space. At least three such productive ludic activities are available for noun-noun compounds. The first invites students to attempt to create the longest possible noun-noun compound, capitalising on the psychological power of competition and an open-ended activity. This can be done in pairs or groups with the students passing a piece of paper between them in turn as each adds a new noun to the emerging compound. Other students can challenge at any time if they feel the resulting compound is too opaque. Figure 8 presents an example beginning with ‘book’.

Figure 9: A lengthy noun-noun compound which at each point may be considered meaningful

Figure 9: A lengthy noun-noun compound which at each point may be considered meaningful

As Figure 9 indicates, new nouns can be added not only at the end, but at the beginning and the middle and indeed a given noun (here ‘investigation’) can recur within the structure, albeit, naturally, referring to different kinds of investigation. The choice of noun is constrained, however, by pragmatic and collocational realities (c.f. the oddness of ‘Yorkshire University camel science textbook yogurt’). A useful twist on this activity is to invite the student who has just added a new noun to the mix to rephrase the string with prepositions or relative clauses (as noted above), thus, ‘Student psychology textbook error investigation’ would be rephrased as ‘an investigation into an error in a textbook for students about psychology’. This invites and requires knowledge of the appropriate prepositions needed to express these meanings.

The second ludic activity follows from the highly productive property of NNCs i.e. that despite the fact that there are pragmatic and collocational constraints to the contexts in which nouns are likely to occur with (an)other noun(s), it remains the case that any two nouns in English can be given a meaning (or indeed several). Figure 10 offers some examples.

Figure 10: entirely novel and semantically ambiguous NNCs

Figure 10: entirely novel and semantically ambiguous NNCs

This activity might be done in pairs of students where each member of the pair writes down a random (singular) noun and the pair discusses possible meanings. This is easily transformed into the ‘pelmanism’ or pairs style of game. Extending this yet further, in triads, three random nouns might be written down and the six resultant dyad possibilities given consideration[14]. A paragraph might then be written using two or three of them. This is exemplified below.

Three base nouns

This activity should serve to highlight the flexibility and creativity of this pattern as well as facilitating growing familiarity with the head-modifier ordering and semantic relationship aspects of NNCs.

A third productive ludic activity for NNCs takes its lead from the Anglo-Saxon game of kenning (Mitchell & Robinson 1992), the use of metaphorical NNCs in poetry and prose to figuratively paraphrase concepts for which a single word already exists. Some examples are given in Figure 12 along with the Germanic literary sources for the kennings.

Figure 12: Some kennings from Germanic literature

Figure 12: Some kennings from Germanic literature

This ancient form of wordplay can be brought readily into the contemporary EFL classroom. Indeed, contemporary English offers its own examples: refrigerator can be rendered as ‘coolbox’; spare bedroom can be rendered as ‘boxroom’. Both these are appropriately metaphorical in the spirit of the Old English kenning game. Having introduced the concept of the kenning, then, this activity would invite students to coin kennings for contemporary phenomena which are typically expressed in only one word. Some possibilities include the following: ‘page journey’ (book); ‘life knot’ (marriage); ‘glue festival’ (wedding); and ‘life tent’ (house). All the above activities are creative and playful whilst at the same time ‘serious’ in their engagement with the linguistic phenomenon of the NNC in English.

2.3. Noun compounds envoi

For reasons of space this article does not consider morphologically complex NCs such as ‘truck driver’ and slum clearance’. However, to round-off this first discussion, one further kind of noun-noun compound in English termed here the appositional compound is briefly discussed. Examples are given in Figure 13.

Figure 13: Appositive compounds in English

Figure 13: Appositive compounds in English

As opposed to NNCs, there is little system or productivity here. However, these patterns are not uncommon lexemes, nor are these entirely unproductive patterns and the ‘appositional’ structure can be extended to non-NNCs e.g. ‘the get-go’. Their inclusion within an English language curriculum is therefore justified. Pedagogically, these are most likely to be approached from a lexical approach point of view (Lewis 1993). A functional syllabus which introduces them with other jobs or with countries might be a way in. Such ‘everyday’ inclusion may normalise the phenomenon of the compound to the student and serve to support the NNCs which this paper argues are introduced more early than has perhaps been typical.

The above section has sketched some awareness-raising and ludic teaching approaches to NNCs, some of which make crucial reference to certain linguistic concepts specifically modifier-head linear order, stress and NNC semantic relationships. The text has addressed the issue of language level namely that NNCs are so prevalent, and form around any kind of word (i.e. are not necessarily ‘academic’ lexis), that they could and should be introduced very early and mentioned one issues that a compounding syllabus might consider i.e. the place of metaphor in the syllabus. It has also considered one relatively peripheral aspect of English compounding, appositives, and argued that a lexical approach (Lewis 1993) within a functional syllabus framework lends itself to these structures. The notion of ludic learning has guided some of the pedagogical suggestions. The second part of this article in the following issue of this journal will consider the pedagogical ramifications of a further type of English compounding, compound adjectives as well as making some more general observations on the interaction of linguistic concepts with language teaching.



  • Figure 1 does not demonstrate all the word formation processes available in human language(s). Some do not occur in English at all for example template morphology (Simpson 2009) which is a feature of Semitic language and infixation. Others are fairly limited in English such as incorporation (Baker 1985) and reduplication. A further type of morphology not listed in the table but present in English is the clitic (Spencer & Luis 2012).
  • This definition of course assumes that syntax does not descend below the level of the word, an assumption rejected by most theories of grammar.
  • Non-referentiality is not always a property of both compound and phrase: ‘a cup of coffee’ contains a particular kind of coffee; a ‘coffee cup’ is a cup for coffee but may not have any coffee in it right now.
  • Occasional exceptions to this rule surface: ‘sports shops’, ‘sportswear’, ‘sports car’; ‘systems analyst’ and (Hewings 2005: 86) ‘savings account’, ‘customs officer’, clothes shop’, ‘arms trade’, ‘glasses case’ and ‘arts festival’’. However, in these compounds, the first element cannot be singular (*’arm trade’); plurality, here, then is fixed unlike in phrasal constructions.
  • A fourth way in which two nouns can be combined which is not discussed in this article is exemplified by the formulation ‘sausage eater’, ‘dragon slayer ‘ and ‘hairdresser’. Although a noun-noun combination these are not NNCs in the sense the term is used here as the second head noun is morphologically complex.
  • This exception may be due to another (sub)rule of how frequent the expression is.
  • We note in passing that the elements in this compound, and those ‘race car’, can undergo inversion and remain meaningful (and used) compounds: ‘car race’; ‘chair arm’.
  • Although the first three elements of this set appear very similar, the semantic relationships may be subtly difference.
  • Singularity is not an inherent feature of these combinations c.f. footnote 5.
  • Exocentric compounds in English also exist (‘pick-pocket’, ‘paperback’) where the head is outside the formal string (i.e. a pick-pocket is a kind of thief; a paperback is a kind of book).
  • Debate continues on the number of semantic relationships and the basis for the classification. See Girju et al. (2005) for a discussion.
  • If the fictional Baptiste character in the introduction is a L1 French speaker (as his name may imply), this explains his error.
  • Cekaite & Aronsson use the term ‘ludic turn’ (p.170) to describe the shift they perceive in methodology towards a respect for the learning power of play. This article adopts this terminology.
  • Of course one could combine more than two nouns two create ‘frog cave cake’. Experience suggests that these are not easy to use in the narrative.


Baker, M. 1985. Incorporation: a theory of grammatical function changing. PhD Thesis. MIT.

Brustad, K., M. Al-Batal & A. Al-Tonsi. 2004. Al-Kitaab Fii TaCallum Al-CArabiyya: A Textbook for Beginning Arabic: Part One (2nd ed.). Georgetown University Press.

Cekaite, A. and Aronsso, K. 2005. Language play, a collaborative resource in children’s L2 learning. Applied Linguistics 26(2), 169-191.

De Castell, S. 2011. Ludic Epistemology: What Game-Based Learning Can Teach Curriculum Studies. Journal of the Canadian Association for Curriculum Studies 8(2)

Girju, R., D. Moldovan, M. Tatu and D. Antohe 2005. On the semantics of noun compounds. Computer Speech and Language 19, 479-496

Hewings, M. 2005. English Grammar in Use, Advanced. Cambridge: CUP.

Huizinga, J. (1950), Homo ludens, The Beacon Press, Boston

Kirkham, D. 2015. Sailing the 7Cs of language learning. IATEFL Voices,

Kolb, A. & D. Kolb 2009. ‘Learning to play, playing to learn: A case study of ludic playing space’. Journal of Organizational Change Management

Lieber, R. & P. Štekauer. 2009. The Oxford handbook of compounding. Oxford: Oxford University Press.

Lewis, M. 1993. The Lexical Approach: the State of ELT and a Way Forward. Hove, Language Teaching Publications.  

Mitchell, B. & F. Robinson 1992. A Guide to Old English. Oxford: Blackwell.

Nastase, V., J. Sayyad-Shirabad, M. Sokolova & S.Szpakowicz. 2006. Learning Noun-Modifier Semantic Relations with Corpus-based and WordNet-based Features. In Proceedings of the 21st National Conference on Artificial Intelligence, pages 781–787. MA: American Association for Artificial Intelligence: Boston.

Lord of War. 2005. [Film]. Andrew Niccol. dir. USA: Saturn Films.

Ryder, M. 1994. Ordered Chaos: the Interpretation of English Noun-Noun Compounds. Berkley: University of California Press.

Schmidt, R. (1990). The role of consciousness in second language learningApplied Linguistics, 11, 129-158.

Simpson, A. 2009. The Origin and Development of Nonconcatenative Morphology. PhD Dissertation, University of California, Berkley

Spencer, A. 1991. Morphological Theory. Oxford: Blackwell.

Spencer, A. & Luis, A. 2012. Clitics: An Introduction. Cambridge: CUP.  

Swan, M. 1995. Practical English Usage (2nd ed.). Oxford: OUP

Widdowson, H.G. 1978. Teaching Language as Communication. London: OUP.

Appendix: List of Learning Resources

The below list details coverage of language materials which can be viewed as teaching compounding in a range of standard coursebooks (Part A) and support materials (Part B). The coursebook resources reviewed have a very similar approach to compounding: it is mentioned very briefly, there is no unpacking of the relevant linguistic concepts, the tasks provided are surface level.

This is less true of the support materials, particularly Quirk et al. (1985), which provides its usually fairly extensive discussion. There is of course, little in the way of exercises, and it is debatable to what extent Quirk et al. is usually referred to as teaching resource. Hewings (2005) and Swan (1995) are both judged acceptable by the current author but in the latter there is again no space for exercises.

Part A1

Part A2

Part B



Deak Kirkham is a postgraduate research student at the University of York, UK, working on the relationship between interactional competence and learner self concept (primarily as a socio-cognitive but also as an existential construct) in adult second language learning. Outside of this focus, he is actively involved in the writing praxis community both as a scholar with presentations / workshops delivered at NAWE (National Association of Writers in Education) and EWCA (European Writing Centres Association) and also, more recently, as a practitioner developing and delivering one- and two-day writing workshops in the community. Deak is also an evangelical Esperantist, active in the UK Esperanto community, with a strong ideological and linguistic commitment to the legacy of L. L. Zamenhof as a creator of a successful constructed language and as a humanistic philosopher.

His undergraduate and postgraduate degrees in Chomskyan generative syntax afford him a further domain of linguistic engagementː the (ir)relevance of linguistic concepts to adult second language learning. While opposing the simplism of ‘standard’ textbook approaches to certain aspects of language learning and the ‘linguistics-lite’ training and professional development of the language teaching profession, Deak finds himself equally perturbed by some current attempts to over-theorise language learning. He is employed at a UK university.


View profile