I thought I heard somewhere that pawns are not pieces but just called pawns. Then, I heard from some other places that they are. I think they are but what is the exact definition of a pawn?
I think a "piece" in the chess context is typically shorthand for the more specific terms of a "minor piece" (Knights and bishops) or a "major piece" (Rooks and Queens).
So while the term is very commonly used and understood to refer to major/minor pieces, I think fundamentally it's at least a little ambiguous and when communicating especially with non-chess players it's probably best to simply be more specific, at least the first time the term is used.
No, the word "piece" does not include pawns. We can see this in some common phrases. A "minor piece" is a bishop or a knight – pawns are even more minor but don't count as minor pieces.
The word "piece" is also used in phrases where it implies "bishop or knight" because if the piece in question had been a queen, rook, or pawn, this would have been specified. For example, when we say that a player is up or down a piece.
One usage that might confuse the issue is that the physical objects we move on the board are called "pieces", and this refers to the pawns as well as to the pieces. We also see this usage when we say that in one game one player "has the White pieces".
Wikipedia lists these usages.
Yes. According to the FIDE Laws of Chess (emphasis mine):
2.2. At the beginning of the game White has 16 light-coloured pieces (the 'white' pieces); Black has 16 dark-coloured pieces (the ‘black’ pieces).
These pieces are as follows:
There are also other articles implying that pawns are pieces:
126.96.36.199 This exchange of a pawn for another piece is called promotion, and the effect of the new piece is immediate.
However, it's worth noting that a pawn isn't considered a piece in terms of the Algebraic System:
C.1 In this description, ‘piece’ means a piece other than a pawn.
I think it's just convention (see appendix). There are some sayings like
'In endgames, when you're winning, trade down. But trade pieces, not pawns.'
If you consider pawns as pieces, then this becomes
'In endgames, when you're winning, trade down. But trade non-pawns instead of pawns.'
Seems pretty weird.
There are also some definitions of the start of endgames like
(for lichess) Endgame starts when there are 6 pieces except kings.
If you consider pawns as pieces, then this becomes.
(for lichess) Endgame starts when there are 6 pieces except kings and pawns.
(I cannot think of examples that do not involve endgames.)
It's like in financial accounting when 'accounts receivable' (basically IOU's) are considered 'tangible' by 'accounting fiction' even when by our everyday understanding of the word 'tangible' they are 'intangible'.
or like the definitions of open and closed in topology: under these definitions, a set can be both 'open' and 'closed' (look up 'clopen'), but in our everyday understanding of these terms a door cannot be both open and closed.
In discussions of strategy, "piece" usually does not include pawns. In some contexts it may just mean knights and bishops. "Black is up a piece and a pawn".
In all other contexts (rules, board setup, physical set design, computer engines, etc) "piece" usually does include pawns.
There are some ambiguous contexts where it's not clear.
For instance, "there were 6 pieces left on the board". If this was a line in a book that is not about chess strategy, I'd assume it includes pawns. If this was a line in a report on a chess match, I'd assume it doesn't (but I wouldn't be certain whether it included the kings).