Tom McArthur’s book is essential reading in the world of reference, although perhaps that is faint praise for a subject that has so few books dedicated to it. However, reading the book is tough going. The author makes it difficult for us to reach the nuggets of genuine insight, since they are often hidden away at the end of chapters that ostensibly are about something completely different. It is as if there are two competing books fighting each other throughout this one. This conflict is revealed in chapter two:

When I first started the work that led to this book I thought I was engaged in outlining the history of lexicography and its related disciplines. … I was in fact toying with a distinct way of looking at human history.

Unfortunately, the author has been as good as his word and has attempted a distinct way of looking at human history (about which more below).

Incidentally, the book is far more than a ‘history of lexicography’. The term ‘lexicography’ has never to my knowledge been used for encyclopedia compilers, whatever the author may say about the interchangeability of the terms ‘encyclopedia’ and ‘dictionary’.

McArthur is a fair if undistinguished compiler of a primer on the history of reference compilation and publishing. He covers a vast span of human history, with the inevitable result that the details can only be sketchy.

What makes the book a challenge to read is that strangely, for a book about reference publishing and the organisation of knowledge, the author’s own organisation is often lacking. Passages seem to drift into areas that have little or no connection with the argument in hand. McArthur seems excessively fond of using etymology to make a point when he must know that the etymology of a word has little or no relevance to its current meaning – or to the theme of the book. For example (page 19):

It is still worth noting … that the term ‘digit’ (so useful in mathematics and computation generally) derives from the Latin word for a finger. (Indeed, it can be argued that we favour decimal arithmetic over duodecimal or binary because we have been blessed with 10 fingers to count on, and not 12 or 2).

What is the relevance of all this in a book about reference?

But McArthur becomes truly interesting when he moves to an area he is familiar with, as when he discusses the two main ways of organising a reference work, thematic and alphabetical, and what led him to compile the Longman Lexicon of Contemporary English (1981). The Lexicon is a truly innovative concept, which provided linked information about a term in a way that a dictionary or encyclopedia cannot do so effectively – for example, that the male equivalent of ‘mare’ is ‘stallion’.

Longman Lexicon 1981

This represents a genuine solution to the problem of A-Z reference works, and McArthur is entitled to dwell on it, regardless of how successful or otherwise the book was when it was published (it didn’t sell well). Of course, today there are digital ways of showing links that print-based publishers could not have imagined. Nobody could read Worlds of Reference today without smiling at McArthur’s excitement at what computers might do for us. The author can be forgiven for being carried away by new technology, but it is a shame that he was not able to bring the argument forward to the digital age, when the distinction between thematic and alphabetical begins to evaporate. However, he seems on much safer ground when he talks about practice rather than theory. His two justifications for thematic arrangement are both based on the idea that thematic rather than A-Z is closer to the way the brain works:

  1. Any reasonably well-constructed conceptual framework is far closer to ‘reality’ and how our minds work than anything that is alphabetically ordered.
  2. There is a considerable consensus … as to what the primary categories need to be.

These are big and unchallenged statements. Just because Francis Bacon and Peter Roget used similar groupings of topics doesn’t mean that (2) is true. Tantalisingly, the illustration on page 151 from a hierarchical lexicon for birds is astonishingly close to a diagram of RDF triples in present-day collections of semantic links. But RDF triples don’t to the best of my knowledge reflect the way the human brain works any more than do alphabetical or thematic arrangements.

McArthur does make some interesting points, albeit in a rather incidental way, for example:

  • The unacknowledged status of reference works as sources of information. We all use Wikipedia today, but how many people cite it?
  • Alphabetical order in publications seems to have developed after the invention of printing. This may suggest how technology affects the way we do things, contradicting his statement on page 167 that ‘the work done [by computers] is the same work as has always been done.’ That’s a very limited view of IT!

Limitations of the work include:

  • Far too many generalisations in the text, drawing grand conclusions from little evidence (for example, in chapter three, he contrasts ‘contingent’ things that were around at the time, such as papyrus, with ‘universal’ things such as ordering and listing. But different societies used different ways of ordering, so I would hesitate to state the ‘ordering’ is universal, valid for all time.
  • Questionable associations between taxonomies and value judgements. In chapter five, he claims that simple classification is hard to achieve without the imposition of special value judgements of better or worse. ‘In modern armies, GENERAL is a superordinate term set above CAPTAIN, SERGEANT, etc … in all such metaphors a higher position indicates a greater capacity, control or ability to contain.’ This is questionable, and hardly relevant.
  • The author accepts publishers’ statements uncritically, and tends to exaggerate the achievement of specific works. Comparing Encyclopaedia Britannica with Diderot’s Encyclopedia, he concludes ‘in the long run, EB has been just as revolutionary’. I’m sure the Britannica publishers would like to think so. McArthur is lavish in his praise of the Oxford English Dictionary (‘the superlatives that have been lavished upon it are in the main well deserved’). The OED is a monumental achievement, but neither innovative (the Grimm brothers started over fifty years earlier) nor particularly radical.

McArthur is also surprisingly simplistic in his view of knowledge. ‘Today we live in complex societies … Once’ he states confidently on page 136, ‘all priests knew much the same thing (with a few specially learned ‘doctors’)’. I have a lovely image of empty-headed priests, all knowing very little, but hopefully happy.