Unlike A.I., booksellers loathe misstatements of fact. Because unlike A.I., booksellers are capable of feeling shame.
But before moving on, I think it’s fair to acknowledge that the antiquarian bookseller’s catalog description is itself an inherently artificial form, designed to convey authority and erudition where, to the degree that either exists, they’re often the product of intense but superficial research. Fragments of that research occasionally linger in the cataloguer’s memory; the accumulated residue of these fragments is valuable in itself, and one thing that makes old booksellers useful. But it’s just as likely that our hastily-acquired expert knowledge will be erased from memory as soon as the book is sold. I’ve catalogued roughly 150,000 books in my career, and put my hands on many more, so I think this is probably a good thing.
In this respect, our process is not so dissimilar from that of generative A.I: we begin by scanning what we judge to be a fair representation of the literature on the topic we’re researching, hoping to find therein some salient nugget that might compel our notional customer to buy our book. We’ll build our process around that nugget, aggregating and polishing until we’ve synthesized a description whose surface suggests to the reader we have long and deep experience with the item at hand, that our judgments are sound and, most of all, that what we’re offering them is worth what we say it is.
Booksellers with long experience in a single subject will of course be faster and better at this process – when dealing with items within their area of specialty, at least – but even the most experienced specialists are regularly presented with items we’ve never seen or knew existed. And the process is the same whether one is a specialist or not. It’s what enables a talented generalist to craft a convincing pitch for an item about which they have no prior knowledge at all.
I’ve tried, but I find I can’t precisely define the impulse that underlies and guides this process. It strikes me as being almost unique to booksellers. A close approximation is: the ability (born of desire? temperament? loneliness?) to anticipate the flicker of interest that crosses a collector’s face when presented with a certain object, and to gratify that flicker of interest with a series of answers to questions they haven’t yet thought to ask.
It’s an ability that shouldn’t be underestimated. When done well and honestly, it’s a powerful tool for discovery, involving not just an understanding of the psychology of the collector, but, far more importantly, I think, an understanding of how literature works: that in any discipline that has produced a written record (I defy you to name one that hasn’t) there exists a hierarchy of texts, ranging from canonical to superfluous to undiscovered; that these categories are fluid, and a text thought canonical in one century may be considered superfluous in another; that the discovery of an unknown text, or of an unremarked facet of a known one, may allow us to successfully argue for a rearrangement of those hierarchies – these are skills booksellers spend their lives honing. We aren’t scholars, most of us; but we know what scholarship looks like, and we live our lives in sometimes unintentional service to scholarship, burying ourselves in dumpsters when necessary in hopes of discovering, if not the next canonical text, then at least something we can flog to tide us over until the real thing appears. No one else does this work – no one else gets paid to do this work – which, though often inglorious, is really valuable work.
Like A.I., booksellers can and do make mistakes. That sheen of authority we project rests on statements of fact, ranging from banal (“372pp”) to more complicated (“first edition, in first issue dustwrapper, first state of the text”) to more complicated still (“the first and still the most-referenced monograph on the Strombidae family of molluscs”) to downright dangerous from a fact-checker’s perspective (“written while the author was on sabbatical from the University of Iowa, between May and November 1972”) – and so on. The more of these facts we present, the likelier we are to be in error about one of them. Because of this, rare book cataloguers have developed a mantra: “it’s better to be vaguely correct than precisely wrong.” The mortal sin, which all (sensible) cataloguers live in fear of committing, is that of asserting something to be true that simply isn’t. We fear it, first of all, because we realize that the minute we’re caught in an obvious misstatement of fact, the entire edifice of our authority begins to crumble. I say “edifice,” but “veneer” is surely the more applicable term in most cases. One puncture and the whole thing hisses away like a bag of gas. Our only armor is to be correct.
But that’s only the self-interested aspect of our concern for truth. If we’re good booksellers we look to a higher power, and that is the credibility and continuity of the trade beyond ourselves. A seemingly innocuous error can do damage far beyond our small reputations. The tendency for mistakes to take on a life of their own and haunt the trade long into the future is a documented thing in the rare book world. When these misstatements are the product of self-interest on the part of a bookseller (i.e., lies), they call the legitimacy of the entire trade into doubt. On this point, I leave you this week with an invitation to read British bookseller Laurence Worms’ terrific essay “Assertive Cataloguing,” written more than a few years ago but more relevant now than ever.
