Harmless drudgery, theft and restoration
When I was a small child, my brother, only a few years older than I, published his first paper in a learned journal. He was an avid butterfly collector and had found a specimen of a species thought to have been extinct for many years in our country. Feelings of family pride were not, I am sad to admit, untinged with unworthy thoughts of sibling rivalry, which I assuaged with reflections on the differences between publishable research in entomology and in the subject which then interested me, mathematics. More than half a century later I have a different passion, one which I have pursued with some intensity for close to half that time, and which has led me closer to butterfly collecting than I should have expected as a child. Numerous metaphors linking pastel and the dust from butterflies’ wings (so easily blown away) can be found in my Florilegium, and go back to Diderot’s salon criticism: indeed, by as early as 1780, the Italian scientist Marsilio Landriani (a founder member of the Milanese Società patriottica appena fondata allo scopo di promuovere le arti, le manifatture, l’agricoltura e le scienze applicate) suggested that butterflies should be preserved using Loriot’s celebrated method for fixing pastels (this led to a discussion in the famous Journal des sçavans).
In his generous introduction to my Dictionary of pastellists before 1800 (2006), Pierre Rosenberg (from another famous academy, and himself a lexicographer by virtue of his own Dictionnaire amoureux du Louvre which I have had occasion to cite before now) called my compilation “un dictionnaire amoureux” – a thought which I have tried to keep in mind during a further eight years of grappling with the good, the bad and the ugly, and of the continuing struggle to overcome our ignorance about artists whose skills continue to take my breath away. From time to time, however, I am puzzled at the process itself – lexicography, albeit in a very specialised sense. And reading the wise reflections on academic research by my friend Luke Drury, president of the Royal Irish Academy, made me want to talk about some of those questions.
Dr Johnson also said that lexicography was “the art or practice of writing dictionaries”, a broad enough definition to include the numerous sightings of extinct pastellists which will readily be granted as my original work, as well as the accumulation and organisation of information known to others. But is a dictionary “just” a compilation, or can a body of knowledge, suitably organised, show signs of emergent behaviour that justify a more scientific status? If so, how does the lexicographer do this? If we understand this process better, can we do it more effectively with different tools?
I shan’t repeat what I’ve already written in an earlier post about connoisseurship and attribution, but I should perhaps rehearse the special difficulties that beset the study of a field so badly neglected as pastel has been for a very long time. Ratouis de Limay’s 1946 book was the last serious attempt – indeed one review of my Dictionary claimed that I had “relied heavily” on it, seemingly unaware that my book covered schools of all nations, had entries on six times as many pastellists, and contained 5000 reproductions against the 100 in Ratouis de Limay. And a lexicographer cannot omit material just because someone else has previously used it, so I don’t apologise for reusing the 2%. In lexicography, plagiarism is not bad manners, but a professional obligation.
But in extending the frontiers so far, the reliability of the discoveries is subject to special hazard. As the great lexicographer H. W. Fowler noted in 1929, “A dictionary-maker, unless he is a monster of omniscience, must deal with a great many matters of which he has no first-hand knowledge.” We would all be grateful to find the steps of other pioneers to tread in, if not shoulders to stand upon.
For one of the responsibilities of the lexicographer (and, dare I confess, one in which the best lexicographers all derive a malicious pleasure covered with the cloak of their intellectual duty) is to find fault with their predecessors. The battle against entropy is conducted on many fronts. But then we are assailed with a different responsibility: as Herodotus puts it (ii:24),
Εἰ δὲ δεἳ μεμψάμενον γνώμας τὰς προκειμένας αὐτὸν περὶ τὣν ἀφανέων γνώμην ἀποδέξασθαι…
(or, loosely, having condemned others’ opinions, I must now say what I think about these obscure matters). The geography of the Nile delta was not more obscure to Herodotus than the work of many pastellists (some quite good) remains to me.
So what are these lexicographical tools, beyond the obvious ingredients of words, to which of course the addition of pictures takes us beyond what was available to Herodotus and even (for colour and number) Ratouis de Limay?
One of the best tips I was given when starting the work was (again from Pierre Rosenberg) not to encumber the text with complex abbreviations. And yet of course the utility of a dictionary is greatly enhanced by compact presentation: no one reads a dictionary, but they turn the pages in search of information in a very specific way that is undermined, perhaps fatally, by poor layout or the inclusion of the wrong kind of information.
Much of the unseen art of the good lexicographer lies in the technique of judicious conflation. Two records describe what may or may not be the same work. When do we recognise them as a single work, when do we introduce a question mark to alert readers to a possible error: even the niceties as to whether the hestitant conflation is marked within the line (i.e. run on) or given a line break all involve a far subtler use of typography than most users need to understand explicitly. If done well, the message is adequately conveyed using common sense rather than express instructions. That is why the online version of the Dictionary uses pdf documents that maintain the appearance of the printed page – something which is instantly lost in a formal structured database (search capability may be enhanced, but at the expense of the subtle gradations I describe). Of course this doesn’t always work: one weakness of my layout is that some people think the Dictionary is a picture book, with captions conventionally placed under the relevant picture, rather than a dictionary, with information (including the image where there is one) under each headword. But this misconception persists despite, in this case, explicit instruction – which just goes to prove why a lexicographer cannot rely on giving complex, counterintuitive rules to readers. Another inevitable corollary of the concision princple is that a week’s work can often result in nothing more than the removal – or addition – of a single question mark.
So, with these blunt instruments, the art-historical lexicographer goes to work, trying to bring to light – if not to life – information that has been lost, garbled or confused. The scalpel of the double question mark is accompanied by the conflationary “=?”, applied in much the same way as the picture restorer uses acetone and watercolour (not on pastels please!). The fundamental principles are always to try to uncover what was there originally, excising the errors of past research, while taking a humble view of our own fallibility by liberal use of reversible suggestions. The same reviewer who thought I had drawn “heavily” from Ratouis de Limay also criticised me for using the term “version” which I defined as “replica or copy”, seemingly (or wilfully) misunderstanding the deliberate ambiguity that is the only proper response where the information is inadequate to make a judgement.
The necessity of compactness has another consequence: much as I should like to acknowledge the assistance I get from time to time from eagle-eyed users, only sources which other users are likely to find helpful appear in the articles: there is no room for the footnotes in which the normal courtesies of scholarship are observed, so that private communications are only cited where they go beyond drawing attention to an existing published source. From time to time I report to the Oxford English Dictionary an earlier occurrence of a word: it is the source I cite rather than my name which I hope to see in future editions. I mention this because I know it can cause Large Blue type feelings, for which I should like to apologise and cover as best I can with a general acknowledgement of all such help – including to the gentleman who sent me an e-mail headed “Mistake in your Dictionary”, telling me he had information about a pastellist which he would only reveal if I agreed to credit him explicitly; he then sent me a reproduction of a chalk drawing made around 1850.
There is another reason for not including names of people with whom I have discussed attributions: unless the full arguments are set out (as they can be in an article, but cannot in a dictionary), it is unfair to reduce often complex, nuanced positions taken by other scholars to the shorthand required.
Since I take a somewhat simplistic view of the value of making information available as widely as possible (subject of course to the obvious concerns of private collectors not to have their ownership disclosed), I have received some complaints. Surprisingly few people have contacted me about attributions they don’t like. It is also easy to see why dealers might have a commercial sensitivity to the publication of provenance where they have bought cheaply at auction: but curiously I am more often asked to remove the auction price (even though any intending purchaser will have no difficulty in establishing this directly) than details of the auction itself. One request however did take me by surprise: where I had reported (with full acknowledgement) a finding by a student which had been presented in public. Apparently having your work cited removes its status as original research for the purposes of one university’s degree requirements. To me such a restriction encourages the wrong sort of behaviour, and transgresses the values of academic research so admirably summarised by Luke Drury.