It just means it gives the opponent a pawn break.
https://chess.stackexchange.com/questions/1181/what-is-a-pawn-break
In other words a way to force lines open by advancing their pawns against the "hook." I guess they call it a hook because it sticks out? Anyway, I've heard the term before, but I don't know its origin.
In contrast if pawns are kept, for lack of a better word, flat, then you can't force lines to open by advancing pawns.
Thank you, Preggo_Basashi .
According to the link :
A pawn break , narrowly defined, involves your pawn blocked by an enemy pawn ('ram' per Kmoch).
You advance an adjacent pawn to the rank of your blocked pawn. This attacks the enemy pawn
( 'lever' Kmoch ). When the enemy pawn captures, it frees your blocked pawn to advance.
White moves c3 or f4 would initiate a pawn break .
However some authors have a broader definition . Consider Soltis Pawn Structure Chess , p 55-56
Black's last move was 5 ...c5 . Soltis describes this as "Black's ...c5 Break"
Getting back to pawn hook . From your answer I assume the initial move to attack a blocking pawn could be referred to as the hook . Two things are still not clear.
The first line of your answer "it gives the opponent a pawn break". To whom does the pawn break
belong ? One would think it was the player who initiated the attacking move.
Finally the book review : Shankland's pitfall list 5. Advancing pawns can become a hook.
It does not make sense that initiating a pawn break is a pitfall. Admittedly I have not yet seen
the book and this line may be "out of context". But assume the player makes a rational decision
to do a pawn break and this is not a blunder. How can starting the break by advancing the pawn be a pitfall ? In this line from the book/review the meaning of hook is still fuzzy.
Thank you, Preggo_Basashi .
According to the link :
A pawn break , narrowly defined, involves your pawn blocked by an enemy pawn ('ram' per Kmoch).
You advance an adjacent pawn to the rank of your blocked pawn. This attacks the enemy pawn
( 'lever' Kmoch ). When the enemy pawn captures, it frees your blocked pawn to advance.
White moves c3 or f4 would initiate a pawn break .
However some authors have a broader definition . Consider Soltis Pawn Structure Chess , p 55-56
Black's last move was 5 ...c5 . Soltis describes this as "Black's ...c5 Break"
Yeah, regardless of the textbook definition, in a practical sense a useful way to think of it is it can't advance... either because it's blocked by a pawn or because advancing it would just lose the pawn.
So Kmoch wants the pawn to be blocked, and Soltis gives an example where it is legal for white to advance the pawn to d5, but doing so would just lose the pawn.
In a real game when your goal is to open lines, this is a useful progression to keep in mind. First control the square in front of the pawn (by physically blocking or otherwise) and secondly bring the pawns into contact with each other (the break).
Getting back to pawn hook . From your answer I assume the initial move to attack a blocking pawn could be referred to as the hook .
As I understand it, in my example the a3 pawn is called the hook (it can't help but get snagged when something moves past it), and the move pawn to b4 by black is the pawn break.
The first line of your answer "it gives the opponent a pawn break". To whom does the pawn break belong ? One would think it was the player who initiated the attacking move.
Yeah, in my example black is the one who has a pawn break at b4. White has a pawn break at b3.
The purpose is to open lines... but open to whose benefit is sometimes tricky business 
In one game vs an expert, where we had both castled kingside, I opened lines against my opponent's king only to discover a few moves later that as a consequence it was my king getting attacked, not his!
(Open lines tend to favor the side who is better developed / has more space / has more pieces in that area)
Finally the book review : Shankland's pitfall list 5. Advancing pawns can become a hook.
It does not make sense that initiating a pawn break is a pitfall. Admittedly I have not yet seen
the book and this line may be "out of context". But assume the player makes a rational decision
to do a pawn break and this is not a blunder. How can starting the break by advancing the pawn be a pitfall ? In this line from the book/review the meaning of hook is still fuzzy.
I gave the link you posted earlier a brief look.
Seems he's talking about this kind of move
Now, you'll probably notice that many popular openings, and many world class players play moves like a3, a6, h3, and h6, sometimes extremely early in the game, and then proceed to castle there right away.
So... yeah, it's not always bad, but it's always doubled edged. When GMs play it they know it cuts both ways and have made their cost benefit analysis. When a beginner plays it, it's usually a waste of time (goes against opening principals), it's usually unnecessary (the opponent's bishop didn't belong on g4 anyway, so h3 was a waste) and not realizing their king is more vulnerable as a result.
With reference to the review of Sam Shankland's book mentioned in post #1, the book's reviewer, in his comments after Black's 7...Rg8! (of the first game fragment analyzed) says....
A hook is “an advanced pawn which can be exploited by the opponent to open lines.”
https://chessbookreviews.wordpress.com/2018/07/03/one-small-step/
Oh, and notice in my example after black plays h6 I say g5 is white's pawn break.
But wait, black can ignore it and push his pawn to h5 right?
Well, yes, but then the f6 knight is lost
This is actually a common idea and useful to keep in mind: instead of blocking the pawn first, the pawn break might "fork" a piece and pawn to achieve the opening of lines.
With reference to the review of Sam Shankland's book mentioned in post #1, the book's reviewer, in his comments after Black's 7...Rg8! (of the first game fragment analyzed) says....
A hook is “an advanced pawn which can be exploited by the opponent to open lines.”
https://chessbookreviews.wordpress.com/2018/07/03/one-small-step/
Oh, so more than pawn breaks the concept would include piece sacrifices to open lines too.
Would Black's typical 3rd move in the Ruy Lopez of 3. a6 count as a pawn hook?
That distinction between "pawn breaks" looked pretty trivial to me.
Yeah, but new people don't know what's out there and they're just trying to get the lay of the land.
Thanks so much for help in understanding these abstruse terms.
imho pawn break might best be used in the narrow sense I mentioned in my second post.
My favorite definition is the simplest. In Neil McDonald's excellent Chess Secrets : The Giants of Strategy he calls it "putting another pawn alongside his most advanced pawn". He traces this idea back to Nimzowitsch. I cannot find "pawn break" in my Quality Chess translation, but Nimzowitsch uses freeing move in his chapter Prophylaxis and the Centre.
In Secrets of Modern Chess Strategy Watson uses the terms "break" & "pawn-break" without explicitly defining them. Sometimes he uses freeing move .
Kotov in The Art of the Middle Game , Chapt 4 is more concerned with the forest than the trees. He classifies the pawn centre as one of five types : closed, open, mobile, fixed or dynamic. The plan of play is chosen according to the type of centre. He describes just advancing the pawn and never uses break.
I note the definition given for hook . "an advanced pawn which can be exploited by the opponent to open lines". I won't touch that one.
"When I use a word - it means just what I choose it to mean ..." Humpty Dumpty
Unfortunately I don't have any copies of Mark Dvoretsky's works. Does anyone know if he attempted a definition of, or even used, pawn break or hook ? Did he use other names for theses ideas ?
Chapter 10 in this excerpt from Sam Shankland's book is all about pawn hooks.....(search "hook")...
http://www.qualitychess.co.uk/ebooks/SmallStepstoGiantImprovement-excerpt.pdf
I haven't seen what others said but hook is a pawn that attacks the enemy pawn structure and forces the opening of lines.
A typical example is Sicilian Dragon and King's Indian defense Saemisch where h-pawn is used as "hook" to catch the big fish!
I don't claim to be an authority on this topic, but from what I've read, it appears that the "hook" is actually the enemy pawn or pawns, which have typically been advanced, and which can then attacked by your pawn(s) to open up lines of attack....for example, see the excerpt of GM Sam Shankland's book, linked in my post#14 (and below), particularly Shankland's comments following White's 9.Qd1! on p.192, where he identifies Black's (the defender) pawns on f5 and g6 as "hooks"...
http://www.qualitychess.co.uk/ebooks/SmallStepstoGiantImprovement-excerpt.pdf
So perhaps a hook pawn can be either an attacker's or defender's pawn? The important point being that it is traded off in order to open lines for the attacker...
Again, as noted in my post #17, GM Sam Shankland (reigning 2018 U.S. Chess Champion) in his recent book (Chapter 10, p.192) refers to the defender's pawn as the hook pawn.
Bryan Smith refers to the attacker's pawn. One of the 2 is wrong.
So far, in this thread at least, I have seen no evidence to definitively conclude that either perspective (i.e., attacker or defender owning the hook pawn) is necessarily right or wrong.
perhaps this explains it
Bryan Smith refers to the attacker's pawn. One of the 2 is wrong.
So far, in this thread at least, I have seen no evidence to definitively conclude that either perspective (i.e., attacker or defender owning the hook pawn) is necessarily right or wrong.
Again , if all pawns that can capture are hook pawns then the term alone is meaningless.
Doesn't seem to me to be necessarily meaningless. It there exist no consensus agreement on who the hook pawn belongs to, then it might simply be considered to be a pawn that can be traded off for attacking and line opening purposes, and whose side it belongs to is simply a matter of the preferred perspective of the annotator, i.e., the attacker's hook pawn vs the defender's hook pawn.
Nevertheless, I have seen the term hook pawn used in several books, and if I remember correctly it has always referred to an advanced pawn of the defender, which is the object of attack (by the attacker) in order to trade it off and open lines (typically a file) for the attacker. So based on this, it appears to me that the chesskid.com article is using the term in an incorrect (or less frequently used) sense. But I am open to strong or conclusive evidence supporting either view.
Maybe this one?
https://www.chesskid.com/article/view/the-h-pawn-hook
Maybe this one?
https://www.chesskid.com/article/view/the-h-pawn-hook
That's the article quoted in post #18.
In a review of Sam Shankland's Small Steps to Giant Improvement : Master Pawn Play in Chess the phrase "advancing pawns can become a hook" is used. Original review is in Chess Life July 2018
http://chessbookreviews.wordpress.com/2018/07/03/one-small-step
"Shankland discusses five typical pitfalls ..." "5. Advancing pawns can become a hook."
I cannot find "hook pawn" or "pawn hook" in Kmoch's Pawn Power in Chess , despite his compulsion to
name everything. Nor is it in Pachman's Modern Chess Strategy , Wolff's book or My System. There are few references to this term on the World Wide Web and no clear explanation/definition.