broadly similar scope and aims will contain comparable stretches of text that show a degree of similarity
unacceptable in other areas of publishing. Off course, some leeway is provided by the fact that dictionaries
come in different shapes and sizes with different markets in mind, and each will carve out its own particular
section of langue appropriate to its own purposes. However, the problem becomes particularly acute when a
publisher sets out to produce a direct competitor to an already existing dictionary in terms of size, number of
headwords, and general focus of coverage.
Given that most working lexicographers continually consult other dictionaries as a matter of routine it is
important that all personnel involved in the dictionarymaking process should be in a position to decide, on
some kind of systematic and uniform basis, what kinds of copying, if any, are acceptable, and which are not.
This is desirable not only in terms of the strict legal requirements of copyright, but also on professional and
ethical grounds. As a freelance lexicographer who has worked for a number of different companies, it is my
suspicion that, at present, no such basis for a common practice exists among publishers of dictionaries in the
UK. This paper will set out the results of my attempts to test that assumption, and it is hoped that this may
provide some pointers towards establishing an agreed set of ground rules determining the extent to which
dictionary publishers may legitimately 'borrow' from each other's publications without fear of litigation or
professional acrimony."
[76] In his discussion of the applicable law in England Williams relies extensively on Copinger et al:
"The vital question remains of what constitutes quality in dictionarymaking. The only commentator who comes near
to tackling this question is Skone James, who includes dictionaries in the category of 'compilations', along with
gazetteers, arithmetical tables, lists of football results, etc. Skone James recognises that similarities between such
works cannot be judged in the same way as similarities between, say, 'original' works of literature. In the case of
such compilations, similarity is not in itself proof of breach of copyright, though it may be used to allege such a
breach. The onus would then be on the defendant to show that this similarity was not the result of copying; evidence
might be adduced from textual detail, or from the circumstances in which the text was produced. A successful defence
might be that the similarity was due to coincidence or 'subconscious copying', though the law is undecided on the
latter.
Page 497 of [2015] 3 All SA 478 (WCC)
The key criterion throughout appears to be whether or not the copier has unfairly appropriated the labour, energies
and resources of the compiler of the source material '. . . the principle is that if one person has, with considerable
labour, compiled the work from various sources which he has digested and arranged, then a defendant who instead of
taking the pains of researching the common sources and obtaining his subject matter from them, simply makes use of
the other's labour and adopts his arrangement perhaps with only slight variations, thus saving himself the pains and
labour which the other used, this will be an illegitimate use.' (Skone James 1991, 187)
The precise implications of these legal principles for working lexicographers are not immediately clear, since none of
the commentators addresses problems that are specific to dictionaries such as words that occur in restricted
environments, words or set phrases with a single standard translation, etc. However, one might hazard a few
tentative conclusions. Firstly, that definitions, examples, translations, etc which 'spring readily to mind', casual
phrases used to fill out examples and standard conventions of presentation would probably not be subject to copyright
protection, whereas, for example, more 'difficult' definitions and translations and more innovative types of
presentation probably would. Secondly, as the above quotation makes clear, that minor variations and otherwise
plagiarised text could not be used to refute allegations of breach of copyright. Thirdly, that the practice routinely
adopted by working lexicographers of having the dictionaries of our competitors open in front of us makes it more
likely that a prosecution for breach of copyright would be successful.
The above remarks, however, probably beg more questions than they answer; the fact that dictionaries do not figure
at all prominently in legal text concerned with copyright is a serious handicap to anyone wishing to formulate a legally
watertight set of guidelines for dictionarymakers."
As I have said, the Williams paper was not intended to contribute to legal writings on the topic, but it does,
nevertheless, neatly sum up the legal and ethical dilemmas confronting lexicographers in the preparation of
new competing works.
[77] Applying the issues raised by Williams by way of a general approach to the case at hand, the sort of
questions that arise include the following:
.
How many of the example sentences in the OUP work (which are claimed to be direct copies of the
Pharos work) are the only plausible or (possible examples) for those headwords?
.
Is the switching of sentence structure, or the substitution of words (be they nouns , adjectives or
adverbs ), in an example sentence a lazy attempt to copy or a deliberate distinction made by the
lexicographer to avoid a complaint of plagiarism?
.
Does the deliberate use of such distinguishing words/phrases deprive the second work of originality?
.
If so, is it within the tolerable limit of reproduction mooted by Williams?
.
Does the use of just one copied phrase in the second work constitute similarity where the headword in
question in that second work has been given a number of other example sentences that are original?
.
How much of the Pharos work was copied from other dictionaries by its compilers, and, therefore, how
much of the Pharos work
Page 498 of [2015] 3 All SA 478 (WCC)
claimed to have been copied by OUP is the product of original thought and industry on the part of
Pharos?.
[78] Our law requires a court determining an alleged copyright infringement to adopt a common sense approach in