ESTC, you complete me.

What on earth, I wonder, took me so long to rediscover the English Short Title Catalogue? Where was my head? It's on-line, free for all to access, and allows the kind of data collection that simply could not be done in an old-timey off-line way. With it one of course runs the risk of becoming a bad statistician, and any conclusions drawn based on findings therein must be heavily qualified; that being said, some of the searches I've conducted have been highly suggestive if not conclusive. Here's a smattering of what I've done in the space of only a few hours:

Search terms: "complete" and "compleat". Exact phrase in title. Any language.

England
Years: # items
1601-1700: 938
1701-1710: 269
1711-1720: 313
1721-1730: 345
1731-1740: 407
1741-1750: 414
1751-1760: 424
1761-1770: 559
1771-1780: 683
1781-1790: 835
1791-1800: 872

England
Years:
1720: 23
1721: 23
1722: 31
1723: 20
1724: 37
1725: 48
1726: 39
1727: 32
1728: 35
1729: 35
1730: 59
1731: 25
1732: 48
1733: 33
1734: 36
1735: 40
1736: 43
1737: 37
1738: 50
1739: 48

Search term: "Dictionary." Exact phrase in title. Any language.

England:

1701-1710: 84
1711-1720: 49
1721-1730: 73
1731-1740: 96
1741-1750: 103
1751-1760: 128
1761-1770: 147
1771-1780: 203
1781-1790: 182
1791-1800: 299

Search Term: "System." Exact word in title. Any language.

England:

1601-1700: 36
1701-1710: 35
1711-1720: 63
1721-1730: 71
1731-1740: 89
1741-1750: 99
1751-1760: 128
1761-1770: 153
1771-1780: 232
1781-1790: 286
1791-1800: 445

(bold indicates largest year-over-year/decade-over-decade number--not %--increase, italics largest decrease)

These numbers will not be exact; though the word may occur in the title, unless one looks through each of them (and we all know what eighteenth-century titles are like--each is in itself near the length of a bible) one won't know precisely how the word is being used. For example, "complete" (or "compleat," which started off the century as the favored spelling but gradually lost ground--it's not until 1761-1770 that there are more "completes" than "compleats") generally refers to one or more of the following:

1. A "complete" collection of an author's works or a bound volume of periodicals
2. Complete as in everything-you-need-to-know; the "complete gardener," "farrier," "gamester," etc. Also in "complete system."
   
3. Complete histories, whether of individuals, nations, events, or subjects, and often including letters, memoirs, declarations, public decrees, acts of government, etc. As in "A Compleat History of Magick," "A Compleat History of the late War in the Netherlands," etc.
   
4. A work including some other complete tool, as in charts, indexes, etc.
   
5. Complete as in perfect, or utmost, as in "complete happiness."
   

The first three are by far the most common, but indicate what to me seem very different qualities of completeness. None, moreover, necessarily guarantees anything of the sort. Setting aside the metaphysical quagmire of the fifth category, let us turn to the others. One could reasonably expect a "complete" collection of Ward's London Spy, for example, which ran for 18 months from 1698-1700, and which was published as an 18 part collection in 1703, to be complete. Nevertheless, the 2nd edition of the collection, published in 1704, is advertised as "much enlarg'd and corrected." Corrected, fine; but enlarged? With what, pray tell? Didn't I buy the complete one? The producers of the Compleat Gardener, likewise, might be happy to add 100 items to its list of herbs from one edition to the next. One might also choose to take issue with the idea that Edward Barnard's New, impartial and complete History of England (covering the period of "earliest authentic information" to 1790) occupies 718 pages in 2⁰, whereas Charles Ashburton's A new and complete History of England (from the first settlement of Brutus to the year 1793) takes up 946 pages, also in 2⁰. Is Ashburton's version of English history somehow 228 pages "more" complete? Some of this is obviously my own naivete; different historians will tell different histories, and perhaps simple word economy could buy one author a couple hundred pages. Seems to me, though, that one man's complete is another's unfinished.

What qualifies an index as complete, by the way, is totally beyond me.

So it would seem that "complete" in the 18th century is somewhat akin to our "All Natural!" Everyone wants it, but it might not mean anything. The numbers of titles including the term in what amounts to a marketing ploy, however, always go up, decade after decade, indicating that (as you'd expect) "completeness," however it was understood, remained a desirable characteristic throughout the century. It is also interesting that the largest jumps in such titles occur from 1760-1790; why then? What was going on in other genres? The epistolary novel, for example, saw a dramatic decline in new titles from 1771-1780, the same decade that saw one of the largest increases in dictionaries, systems, and those miscellaneous genres in which "completeness" could reasonably be advertised ("complete" frequently modified "dictionary" and "system" as well, though one perhaps incorrectly feels that "system" implies completeness--isn't an incomplete system a system that isn't a system?). Perhaps these are related; perhaps not.

What can we conclude, if anything, about the year-over-year changes from 1729-1733? All was holding steady until '29 (a year after Chambers' Cyclopaedia), then in 1730 we have something of a glut. Did booksellers respond by turning down titles over the next year? Did they then think they'd gone too far and respond by jacking up the number again?

I don't wish to fall into the Franco Moretti track of dubious quantification built largely on even more dubious generic classifications (a la Graphs, Maps, and Trees). Taxonomy in the arts, and particularly in the realm of eighteenth-century literature, always seems to end in tears; a picaresque isn't necessarily only a picaresque, and a novel in the later sense of the term could easily incorporate, interpolate, or sublimate what was once understood to be romance or indeed anything else (those of you familiar with my hobbyhorse will know that in my estimation the designation "novel" for much the period is dependent on this ability).

So if one is to use this sort of data, it has to be done carefully. The strange thing seems to me to be how compelling I find it--how willing I am to be convinced by a mere display of numbers. Mary Poovey tells me--or rather, A History of the Modern Fact tells me--that I owe this allegiance to numerical facticity largely to the same period that saw the emergence of the novel (I have ideas about that but they're for another time). There's something very sneaky about literary statistics. 445 items with "system" in the title, with probably just under 400 of them reflecting what I understand a system in the period to be. I run up here against the same old problem--with 445 to look through, in a single decade, and the inability to manage effectively the perhaps over 1000 distinct titles published throughout the century, how can I ever really be sure of what I'm seeing?

Soon--a bit more of the same on the romance and the novel...