Skip to navigation – Site map

HomeSupplementsBook reviews2009Thomas Herbst, David Heath, Ian F...


Thomas Herbst, David Heath, Ian F. Roe, et al., A Valency Dictionary of English: a corpus-based analysis of the complementation patterns of English verbs, nouns and adjectives

Mouton de Gruyter, 2004, 961 pages
Nicolas Ballier
Bibliographical reference

Thomas Herbst, David Heath, Ian F. Roe

A Valency Dictionary of English: a corpus-based analysis of the complementation patterns of English verbs, nouns and adjectives. Mouton de Gruyter, 2004. ISBN : 3-11-017194-5, Prix : 198 €, 961 pages

Full text


1This dictionary provides a valency description of English consisting of 511 verbs, 274 nouns and 544 adjectives. The selection is justified (xl) on the basis of interesting but potentially conflicting criteria: “frequency, complexity of valency structures and potential difficulty for the foreign learner”. Though the last criterion deserves praise as it meets a real demand for EFL learners in terms of competing patterns, the convergence of the three criteria could be re-examined. Some constructions may be difficult for EFL learners solely because they are different in their native languages, not least because the most frequent conflicting valencies on different languages are often for cognates. There is no entry for preside, whereas French, Italian and Spanish have different constructions than English in this case. The most frequent patterns may not be described for advanced learners (a standpoint adopted for example in the bilingual English-French dictionary Le Petit Ophrys where common nouns are omitted). Even so, the dictionary offers the most complex cases of frequently occurring lexical items, and that is undeniably of the utmost importance to EFL learners.

2In this conception of valency, subjects count as a kind of complement, hence the recognition of a tetra- or quadrivalent active use, that is a pattern with four complements as in Someone had suggested it to him as an investment. The valency description comprises statements about the quantitative valency of the lexical units established, an inventory of their obligatory, contextually optional, and purely optional, complements as well as systematic information on the semantic and collocational properties of the complements. An outline of the model of valency theory used in this dictionary is provided in the introduction. This review will tackle some of the points, such as frequency, sampling, semantic comments and metelexicographic insight.

Frequency estimates

3The frequency guidelines are useful though not systematic. Strict percentages (eg “more than 30 %”) are rare. The usefulness of the distinction frequent vs. very frequent might be questioned, as well as the relevance of the label “rare” in some cases. The notion that the construction “tendency that” should be “rare” (862) seems dubious to the present reviewer. That being said, the example in context (“There’s always the strong tendency that I’m going to fall asleep or, even worse, not be able to perform properly.”) provides a relevant collocational adjective (176 occurrences in Google) though “central tendency” sounds much more promising. Some other EFL dictionaries (like the modestly acronymed MEDAL dictionary) distinguish three levels of frequency but they do so with tokens, not with types of constructions, which is a much more challenging issue.

4The number of nouns appears to the reviewer to be slightly underestimated in comparison, as evidenced by their only accounting for 20.6 % of this 1,329-entry dictionary. Since the number of nouns is relatively small, a recapitulative table of the different constructions would have been helpful, as is the case with the appendix of Schmid 2000. The consequence is that there are no entries for the letters X and Y and only zeal for Z. As to W, the list of entries reads as follows: wait (V), wake (V), walk (V), wander (V), want (V), war (N), warn (V), wary (adj), wash (V), waste (V), watch (V), weak (adj), wear (V), weary (adj), weigh (V), welcome (V), well-informed (adj), well-known (adj), well-off (adj), whisper (V), whistle (V), wicked (adj), wild (adj), will (N), win (N), win (V), wise(adj), wish (V), witness (N), wonder (V), word (N), work (N), work (V), worry (V), worth (adj), worthy (adj), wrap (V), write (V), wrong (adj). This short list is almost self-explanatory in terms of issues still pending: compounds, conversions and derived forms might appear to be somewhat under-represented, though the complexity of the different patterns was a criterion for inclusion. The unfortunate consequence is the common absence of the derived noun (commit V, but no commitment) so that potentially diverging patterns between nouns and verbs cannot always be systematically compared in the dictionary.

Noun governing that-clauses

5For each of the 1,329 entries taken into account in the dictionary, a list of the complementation patterns given as “comprehensive” was established with the COBUILD/Birmingham corpus. Though some of the occurrences in his ongoing corpus of that clause complement-taking nouns would probably be appropriately deemed rare, this reviewer believes that some nouns could occur with the N-that constructions. CLUE would be a case in point, as well as JOKE, GIFT, HABIT, TRICK and TROUBLE.

6In the case of noun complement clauses, the reviewer appreciated the distinction made between a comment from / a comment by, in line with stance-taking strategies expressed in the PP, but wondered whether the genitive may not require the introduction of another pattern / valency requirement. Fault or comment are labelled as [it] + that CL but this seems to ignore the in these two cases, as evidenced in the example given under fault (It was not entirely Edward’s fault that he never quite fulfilled expectations). In the analysis of valency for nouns, the genitive could be considered for inclusion, just like the by-phrases are (see insistence by developed countries that) as in this example of insight (Reagan’s book confirms Christopher Matthews’s insight that Reagan’s calling in life was not as an actor but as an announcer, 432)


7Interestingly enough, in spite of the all-encompassing Bank of England that was used for the dictionary, the need was nevertheless felt to resort to some invented examples (but rightfully signalled as such) as with He closed on an excellent 67 after 18 holes (143). This pattern also exemplifies a pattern likely to occur in the press, the SCORE construction, an innovative but realistic assumption that numbers play such a role in newspapers (sport) that they ought to be signalled as such in the guise of specific labels.

Metalexicographic perspective

8From a metalexicographic angle, the dictionary bears testimony to the possibility of maintaining a high standard of values while proposing numerous innovations in the description of patterns. Among the many innovations, the distinction is indicated for many complements between active and passive constructions with the A / P distinction:

A complement can occur as subject of a finite active clause (if no patterns are specified, this complement can act as subject in all patterns of that sense),
P complement can occur as subject of a finite passive clause,
A/P:it when occurring as subject, extraposition with a dummy subject
it is obligatory,
A/P(it) when occurring as subject, extraposition with a dummy subject
it is possible,
(though such a general scheme may tend to override extraposition as a syntactic feature).

9The interest in passive and active feature may have led to an underestimation of the importance of extraposition, since some extraposition features are listed under active/passive. The boundary is not always clear-cut, as evidenced by the remark about PRAY (605) + for that – CLP (only passive) as in That rain should come was prayed for by the congregation. In this (invented) example, the passive form affects the verb, not the that-clause, which the label seems to imply.

10Among other innovative symbols is the indication of alternative word orders, for example <--> links two elements of a valency pattern where the order of the elements is not fixed.

11Even so, special status is granted for “particle” word order: <-> only in pattern + N + ADV <->. In other words the ADV-complement can be realised by a particle, i.e. an adverb such as in, over, up, which can precede or follow the N-complement, for example:

Oily waves washed ashore seaweed and general decomposition.
Oily waves
washed seaweed and general decomposition ashore.

12Semantically, the dictionary offers subtle labels such as [ADV:QUALITY] (the car rides well) for middle voice constructions or the label [DESCRIPTION] for autonymic uses of verbs (the file marked “divorce”). Among the distinctions that may appeal to utterance-based grammarians, are that-omitting constructions and the taking into account of to as a specific marker. Semantic categories reveal the quest for ever more finely grained distinctions of patterns. The relevance of quotatives is acknowledged in the dictionary, as with explode (D2, 291). The relevance of the indirect / direct speech distinction is maintained in the labels QUOTE / SENTENCE. Though quotatives should definitely be included (for uses of the verb like he offered, he exploded), it might be debatable whether this stylistic distinction need be maintained, especially since most entries appear with QUOTE / SENTENCE as a mixed label.

13Some metalexicographic subclassifications may have to be substantiated, especially for corpus-based studies. The ontology (and typology) of patterns is not necessarily established beyond all argument and some choices may appear to be debatable. At least the dictionary does give food for thought in the classification of patterns (whether it be p. 300 for the seven alternative patterns acknowledged for fall behind / against or in the apparent subcategorisation of God as a specific NP for the verb pray, [N:God]P). It begs the question as to whether some patterns are really valency-ruled, and lexically constrained by the selectional properties of the lexical units concerned. This is the case for NEED, where the extraposed form with premodifier is listed as being a specific construction:

P3 [it] + [with premodifier] + that-CL (rare) It’s almost a psychological need that I have to have my car with me.

14The question of the rarity of such a pattern needs to be raised, in the light of the existence of the need + that-CL construction (unattested in the dictionary but found in the BNC in “But I have yet to be convinced of the need that these should be taken at two particular times in a student’s school career”, BNC, ASY: 1338). Such a valency pattern seems to postulate that the pattern is lexically constrained and that extrapositions are not plausible otherwise, though the need that I have to have my car with me is almost psychological would sound acceptable to me. It also seems to imply that only NEED may enable such a construction, which I think is dubious. The underlying question is whether we should consider these “valency patterns” as being lexically constrained or grammar-ruled. In that sense, because of its finer-grained valency patterns, the dictionary clearly tends to treat extraposition as lexical.

Semantic glosses

15Even though no systematic treatment of aspectual distinctions (processes / result) is offered, a laudable attempt to recap polysemy by means of the valency classification (three main meanings for distinction) is proposed for many entries. Among useful tips for EFL learners, specifications are given about classic differences such as between economic vs. economical (262).

16The overall project of the dictionary has far-reaching implications as to the description of co-occurrence restrictions, maybe another name for syntax on a lexico-semantic basis. A practical as well as a theoretical consequence for EFL learners is the existence of competing patterns. Hesitations (familiar to / familiar with) or the question still pending of competing constructions are not finally settled as far as foreign learners are concerned, since it is acknowledged that competing patterns may exist without any apparent semantic distinction. A case in point is given on p. 265: A person can be elated about, at, by, over or with some event that pleased them very much - a useful reminder that semantic classifications such as patient / theme /source may vary subjectively because of the governing preposition. Yet, the question remains, should we not acknowledge a specific semantic interpretation for each preposition governed (see also insight into, about or ona complex situation is ‘a full understanding of it or a statement that reflects such understanding’, 432)?

17As to the shortcomings or rather shortcuts of some of the semantic glosses, I will just pinpoint one of the interesting presuppositions contained in the definition of engaged. If we are to take the following distinctions as being mutually exclusive, there are progressive inferences to be drawn from this semantic characterisation:

Two people who are engaged have promised to marry each other. A person can be engaged
(i) to be married
(ii) to a person of the opposite sex.


18Attention is paid to collocations, not only in the choice of examples, but also with sub-entries. The dictionary makes strong points about the relevance of collocational constructions and proposes undeniably fascinating cases such as drive home (to be compared with press home or hit home). We can certainly take advantage of these examples to further our understanding of such structures, as the multiplication of agency patterns/constructions provides sound evidence of lexicalisation in the making.


19In spite of these (inevitable) shortcomings, this dictionary represents an undeniable breakthrough in the field of corpus-based dictionaries. When they say in the preface, “We hope the combination of corpus research and the theoretical background of valency theory has resulted in a dictionary that the envisaged users find as useful as we hope they will”, this goal has undoubtedly been achieved. This 12-year project is an achievement that will be invaluable for English teachers and researchers alike. Despite some conspicuous omissions, this dictionary will soon make a place for itself as a reference tool in translation studies for non-native speakers of English, alongside the BBI Combinatory Dictionary of English Word Combinations.

Top of page


Benson Morton, Benson Evelyn & Ilson Robert, The BBI Combinatory Dictionary of English Word Combinations, Benjamins, 1986.

Paillard Michel et al., Le Petit Ophrys, dictionnaire français/anglais, Ophrys, 2000.

Schmid Hans-Jorg, English Abstract Nouns as Conceptual Shells. From Corpus to Cognition, The Hague, Mouton de Gruyter, 2000.

Top of page


Electronic reference

Nicolas Ballier, Thomas Herbst, David Heath, Ian F. Roe, et al., A Valency Dictionary of English: a corpus-based analysis of the complementation patterns of English verbs, nouns and adjectivesLexis [Online], Book reviews, Online since 10 May 2009, connection on 19 May 2021. URL:; DOI:

Top of page

About the author

Nicolas Ballier

Nicolas Ballier, Department of English, UFR LSHS, University Paris 13, France.

Pr. Nicolas Ballier has taught English linguistics since 1992 at the Universities of Limoges, Rouen and Paris 13. He is in charge of the teaching of the bilingual corpora module for the master pro TILDE (a vocationally-oriented Master’s Degree about Digital and Linguistic Processing of texts) and teaches linguistics in undergraduate courses and MA seminars. He belongs to a Contrastive Linguistics research lab (CRIDAF, EA 453) and to various linguistic societies. He is the main convenor and series editor of the April English phonology conference at Paris 13. Besides English phonology, his research interests include the linguistics of interfaces and the syntax of noun complement clauses in English.

Top of page
Search OpenEdition Search

You will be redirected to OpenEdition Search