Dare to Question?
Dermot Rochford, June 2023
Over the last few years researchers/chess collectors have been discussing bishop versus knight designs in early (18th century and before) French playing chess sets. The question being, which chess piece in the diagram displayed in Diderot’s Encyclopédie book of 1771, represented the bishop and which represented the knight. See image below Ref. ‘A’
Ref.’A’
In Diderot’s accompanying text it states that the cavalier (horseman/knight) is the 3rd piece from the left. And that the fou (fool/bishop) is the 4th piece from the left (slightly shorter) of the two pieces in the red box above. See text below.
A number of experienced chess collectors have argued that a ‘typo’ had occurred in the text and that the identification of the knight and bishop should be reversed. They have supported their view with well researched and detailed arguments which can be seen in various articles and letters published in editions of the CCI Magazines over the last 3 to 4 years. The other side of the argument, that there is no factual evidence that an error occurred in Diderot’s text and therefore we should be prepared to accept the published text at face value.
Recently when I was working on an article on 18th Century French chess set designs (see link below) some issues came to mind about the ‘typo’ theory.
www.chessantiquesonline.com/frenchdesign/FrenchSetsDesigns.html
This is because in our website article, I had shown that an almost universal principle of knight chessmen being taller than bishops had applied in pre and post Revolution (1789) French playing chess sets. I had called this the Taller Knights Principle (TKnP). So could this principle also be present in the diagram of ‘Encyclopédie’ pieces in Diderot’s publication?
If this ‘TKnP’ principle did indeed apply, then it would tend to support the case that the Diderot text was correct (and there was no 'typo' error) and that the placement order of the pieces in the diagram, while not conforming to modern layout (King, Queen, Bishop, Knight, Rook and Pawn from left to right), was not contravening any standard layout practice of that period.
Fellow CCI French member, Patrice Plain’s diagram (Ref. ‘B’ below) is a very useful starting point in trying to differentiate between bishops and knight pieces in abstract (i.e. non-figural knights) French chess sets.
Ref.’B’
This diagram demonstrates that the top of the bishop piece can have a ‘cutaway’ or a ‘slots’ format while the knight has a tricorn type top.
Variations on these design tops included bishops having tops with a combination of the ‘cutaway’ and ‘slots’ or indeed with 2 ‘cutaways’.
These bishop tops variations can be more angular in shape and be confused with tricorn tops which some commentators previously stated were exclusively associated with knight pieces in early French playing sets. More on this later in the article.
From my examination of my own sets and chess catalog images of French sets, it seems to me that, almost invariably, where tops of bishop pieces have 2 ‘cutaways’ there is a clear semi-circle shape to the back portion of the top which is sometimes punctuated with ‘slots’.
This point is illustrated in image ‘C’ below which shows a selection of bishop top variations from French sets - the three on the left are from sets with figural knights’ heads (so we can be certain the pieces displayed represent bishops) and the four on the right are from abstract type sets.
Ref.’C’
So from this image it can be seen that in all the bishop pieces having 2 side cutaways, there is also a definite semi-circle present in the top.
Bearing this in mind, and reviewing the knights and bishops from an incomplete 'Encyclopédie' type set (image Ref. ‘D’ below) shows that the pieces with its 2 cutaways and slot/teeth must be the bishops and so the other taller piece (other than the king and queen pieces) must be the knight.
This group of pieces is plain in construction and may be earlier in date than the design in Diderot’s 1771 publication.
Ref.’D’
Also on our site, there is an even earlier example of an 'Encyclopédie' type set (P208) in image Ref ‘E’ below.
Ref 'E’
The third piece from the left in this example has the 2 side cutaways and semi-circle type top and therefore this represents the bishop. The taller 4th piece from the left is the knight and so this is another example of an 'Encyclopédie' type design chess set conforming with the 'TKnP'.
An 18th Century ‘Encyclopédie’ set owned by fellow collector, Peter Armit, is another example of an early French set having knights which are taller than the bishops. See image ‘F’ below
Ref. ‘F’
These three examples above, demonstrate the 'TKnP' does apply in at least some of the early 'Encyclopédie' type design chess sets.
Another piece of new evidence comes from a French set which we recently acquired and is dated 1765. This set (which was in Jim Joannou's article in the last edition of the CCI Magazine - Spring 2023), has ‘Dragoon’ helmeted knights which are taller than the bishops as shown in image ‘G’ below.
Ref.’G’
Given this is a chess set dated to a period prior to the publication of the Encyclopédie in 1771, it is the most concrete evidence that the 'TKnP' existed in earlier chess sets. Most importantly, it also proves that a bishop could be represented by a triangle type top in early sets (i.e. pre ‘Encyclopédie’) .
Three further pieces of evidence supporting the ‘TKnP’ should be considered. The first is from Schach Partie, page 217 which shows an image (Ref ’H’ below) of an 'Encyclopédie' set with the asymmetrical topped piece in the bishop position.
Ref 'H'
The second is from Spielwelten der Kunst book, page 93, which shows a mid 18th century French Regence type set (Nr. 13) where the bishop with slot cuts, is smaller than the knight - see image (Ref ’I’ below) '.
Ref.’I’
The third is from Christie’s chess catalog of May 2007, Lot 61 - this also also shows an 'Encyclopédie' set where the bishop is described as having assymetrical knop finials (see image Ref.’J’ below) and this piece also is smaller than the knight.
Ref. 'J'
It is also worth noting:
These same details can be seen in Diderot’s diagram in image ‘A’ above.
Finally, in a post Diderot French book, Hocquart’s ‘Elemens du Jeu des Echets’ published in 1810, a diagram (see Ref. ‘K’ below) was shown of a ‘Directoire’ type chess set with six pieces in descending height order but the text nominated the 3rd tallest piece as the cavalier (knight) and the 4th tallest as the fou (bishop) - therefore conforming to the ‘TKnP’.
Ref.’K’
I am aware that it has been argued that the titles given to the knight and bishop pieces in this diagram resulted from a ‘copy-cat’ mistake taken from Diderot’s earlier diagram and accompanying text. However bearing in mind that:
In my opinion makes the proposition that a ‘copy-cat’ mistake occurred, a lot less likely.
Furthermore it should be noted that in this diagram (Ref.’K’), the queen, bishop and pawn all have similar tops - a regular feature of late 18th /early 19th century French chess sets - and this therefore is further evidence that it is unlikely that there was a mistake or ‘typo’ in the identification of the pieces in Hocquart’s diagram.
So from all of the above evidence I think it is safe to say ‘TknP’ applies, to a very dominant degree, in pre-revolutionary French playing sets of all designs, including the ‘Encyclopédie’ design sets.
One other important piece of information, which emerged from my research, concerns the question of what evidence exists to support the assertion that tricorn form tops were mainly associated with the knight piece in early French playing sets. I can say that, outside the disputed ‘Encyclopédie’ design sets, I found no evidence of an early French playing set, having a knight with a tricorn type top.
On the other hand, I found two other examples (in addition to the set in image ‘G’ above) of chess sets which have bishops with tricorn type tops.
The first is a French set from circa 1760 in the collection of Andreas Wittischek which has a distinctive tricorn form to the bishops - see image ref L– below.
‘Ref L’
The second comes from Christies’ catalog of May 2007 where, in Lot 41’ there is a Directoire type French set c. 1800 with a tricorn type top to the bishop, see ref M below.
Ref ‘M’
So on the balance of probability alone, these three examples above suggest that the chess piece with the tricorn top, shown in Diderot’s diagram, is the bishop piece, just as he stated in his accompanying text.
Summary
In this article, I have not attempted to analyze or counter the specific arguments put forward in support of the ‘typo’ theory. Instead I have concentrated on presenting a range of evidence which to me suggests there is, at a minimum, a strong case to be made that Diderot’s text describing the chess pieces in the diagram, may not have contained a printer’s error and was exactly as he intended it to be.
I do recognise it is indeed a complex matter and I hope that the evidence presented above, shows that French set patterns of the 18th century ( including ‘Encyclopédie’ sets) seem to conform to the ‘TknP’, should have some bearing on the ongoing debate on the issue. This and the fact that most evidence suggests that, in early French playing sets, it was the bishop and not the knight that was likely to have tricorn tops, means that it makes sense for the chess collecting community to try to understand more fully as to why the diagram example of an 'Encyclopédie' chess set shown in Diderot's publication of 1771, could be the exception to these norms. Furthermore, the fact that there is no evidence (as far as I am aware) that Diderot or any of his collaborators ever issued an 'erratum' confirming a ‘typo’ had occurred re the text associated with the chess pieces diagram, only adds to the concern that, relying now (250 years after the event) on the somewhat speculative 'typo' error theory to overrule what was actually written by Diderot, may not be totally sound.
Conclusion
So I believe there is a strong case to be made that we should be prepared to ‘keep an open mind’ on this question and hopefully more factual evidence (akin to that which has recently become available with the dated chess set, as per Ref. ‘G’ above) will surface in the future and this will help resolve the issue in a satisfactory manner.
---------------------------------------------
