Thursday, June 22, 2017

Who Made These Marks?

Let me note quietly the sixth anniversary of this blog!

There is an odd passage on pages 35-6 of Whist.4, the fourth edition of Hoyle's Short Treatise on Whist (1743). Hoyle gives one of his "cases" in a chapter called "Particular Games, and the Manner in which they are to be played...":
Suppose you have Ace, Queen, and three small trumps; Ace, Queen, Ten, and Nine of another Suit; with two small cards of each of the other suits. Your partner leads to your Ace, Knave, Ten, and Nine; and as the game requires rather to deceive your adversaries, than to inform your partner, put up the nine, which naturally leads the adversary to play trumps, if he wins that card. As soon as trumps are played to you, return them upon your adversary, keeping the command in your own hand. If your adversary who led trumps to you, puts up a trump which your partner cannot win, if he has no good suit of his own to play, he will return your partner's lead, imaging that suit lies between his partner and yours; if this finesse of yours should succeed, you will be a great gainer by it, but scarcely possible to be a loser.
That paragraph should give you a good sense of Hoyle's prose. The oddity is that your side suit is given as AQT9, but your partner is said to lead to your AJT9--an obvious inconsistency in the text.

I have three copies of the book and in two of them, there is a hand correction in pencil, making the text consistent by changing AJT9 to AQT9:

Whist.4 (copy 1)

Whist.4 (copy 2)

I have a third copy which is uncorrected. Similarly, the copy in the Bodleian Library digitized in ECCO is uncorrected. But here is one more copy:

Whist.4 (Fox copy)

Interesting! Here it is the holding of AQT9 that is corrected to AJT9, in ink rather than pencil. I know of another five copies, but don't know what appears on page 35--that's probably worth some emails.

How did the error come to be?  That's actually easy to answer. Whist.4 was printed by the same printer who was responsible for the earlier third edition, Whist.3. Indeed some of the type was left standing between Whist.3 (advertised March 18, 1743) and Whist.4 (June 29). The rest is a line-for-line resetting. Let's look at the section from Whist.3:


You can see that the line breaks are identical to Whist.4, consistent with a line-for-line resetting. But here, the hand is AQxxx AJT9 xx xx and partner leads to your AJT9. So, there is no error in Whist.3, nor in Whist.1 or Whist.2 for that matter. Clearly this was an error made by the printer in resetting the type.

Who do you think made the various corrections? It could be either Diligent Readers or a Contrite Printer. In favor of the Diligent Reader view is that the manner of correction differs in the three copies, suggesting different people made the correction. Second, only one of the corrections matches the text from Whist.3, text that would have been available to a Contrite Printer, but likely not to readers. In favor of the Contrite Printer view is that there are other minor errors in Whist.4 that, as far as I can tell, no reader has been diligent enough to correct.

What I do know is that the error persists for a long time. Whist.5 (1744), Whist.6 (1746), and Whist.7.1 (1747) all contain the error. When Thomas Osborne ceased publishing individual works, including whist and other games in Mr. Hoyle's Treatises (discussed here), the error continues until 1763 in Games.4. And when it was corrected, it didn't match Hoyle's original text!

Games.4 (1763)

These later editions suggest another argument for the Contrite Printer: I have never seen the error corrected in any of the later editions. In the end, I just can't say who made the corrections in Whist.4. It would be helpful, I suppose to look at more copies. This is another instance of reward from following the bibliographer's mantra, examine as many copies of a book as possible.

If someone plans to do a critical edition of Hoyle's Whist, this is one of the many little trouble spots!