Because I'm not Saemisch or Spielmann, much less Kasparov or Carlsen, I often make moves that offer material for what I hope will be ultimate advantage without having calculated the outcome in detail. I prefer to think of these as sacrifices, even if (by good luck) they turn out to be totally sound, as shown afterward by detailed analysis.

I think the main issue is one of "terminology theory". We define terms all the time. If we do a good job defining terms, those terms will make it possible to clearly express an idea in a natural way. A poor selection of terms will be awkward and confusing.

In this case, we are struggling with a poor choice of chess terms. We have "sham sacrifice" and "true sacrifice", which leaves the base term "sacrifice" dangling with no clear meaning. First, one should clearly define the base term, and then add qualifiers if necessary.

Also, I dislike the qualifier "true". If a sacrifice is not a "true sacrifice", what is it? Is it a wrong to call it a sacrifice at all?

Finally, I think that when the definition of a term depends on subjective ideas, it greatly reduces the communication value of the term. The definitions of "true sacrifice" and "sham sacrifice" are loaded with subjective ideas. Worse, the subjectiveness belongs to the players, not to us observers. So, we really have no right to be throwing these subjective terms around.

~~~~~

All that to say this: I don't use the these terms because I don't want to propogate poorly defined terminology. I prefer to define "sacrifice" as any deliberate surrendering of material to achieve some purpose. It's simple and unambiguous.

I saw a similar argument a couple of years ago on an aviation website. One group quibbled with the english habit of referring to a concorde as just "Concorde" as if it were a proper name. This group argued that we don't talk about other aircraft in this way, so we shouldn't make any distinction for Concorde.

Ranged against this group were the Brits who argued that Concorde was called concorde, and that was that.

This argument got very heated and unpleasant, with people who had previously been good friends insulting each other (in similar terms to those that have been used agaist me recently). And no-one really understood why it had got so heated over such a small issue. Eventually the mods had to step in, close the thread and ban any further discussion on the subject.

What was really happening was a clash of cultures and different treatments of language. American English is generally rule-based, so the American posters could not understand how Concorde could be treated differently from other aircraft. By contrast, UK English is precedent-based, so under this way of thinking Concorde is called concorde because it just is.

Another example of this is that American pronunciation is generally phonetic, but UK pronunciation is based on custom and usage with many exceptions to the rules. So lieutenant is pronounced "leftenant" in the UK and "lootenant" in the US.

<Once> was being kind and gentle when he wrote < One school of thought is that language ought to be fixed by a standard definition; but another viewpoint is that language changes and evolves so standard definitions can only be a guide.>

Well, I agree with all of this, and also with what <Once> said about the nature and evolution of language.

I admit that my background is not in linguistics, but in mathematics and software development. I know that in these fields it behooves us to define terms that can be clearly understood, and that are distinct from other terms in ways (and only in ways) that are meaningful. If not, it makes communication error-prone and inefficient -- often resulting in arguments of semantics rather than the subject actually in question.

As you say, fuzziness and ambiguity are pervasive in natural language. Still, when we have an opportunity to define new terminology, it is to our advantage to give it some careful thought such that fuzziness and ambiguity are minimized, no?

<YouRang> Fine ideas, in principle. In practice, it's actually surprising how few deliberate constructions survive in the language. Most changes seem to come from slang, euphemism, and sound-changes which are themselves caused by everything from laziness to over-scrupulous care for correctness.

I find it's useful to pause occasionally and remember that every single word we use today was once a 'nasty' novelty -- slang, jargon, slipshod pronunciation, or whatever.

In this specific case, the word 'sacrifice' has connotations that vary between speakers. Any attempt to define it precisely will be influenced by these connotations, whether the definers know it or not.

For my own purposes, however, I shall still avoid terms like "true sacrifice" vs. "sham sacrifice", and my use of the word "sacrifice" (in chess) will be based on the precise and unambiguous definition I offered above.

< the precise and unambiguous definition I offered above.> And on *my* screen, 'above' is exactly where it is. But have you noticed how some people say 'above' (meaning 'below') and vice versa?

Anyway, I understand and agree with what you are saying about languages and how they are ambiguous and constantly evolving.

However, I still think there is such a thing as a poorly defined term and a well defined term. The difference between the two may be relative and temporary, but communication tends to flow better when using the terms that are well defined at the time.

One way to tell if a term is well defined or poorly defined is this: If it's well defined, then most people we easily agree to adopt that definition and use it. If it's poorly defined, then many people will resist using it and argue about what it really means -- such as what we have seen here.

This takes me back to my point: The way some people want to define 'sacrifice' (in chess) appears to be poorly defined simply because it prompts resistance and debate.

For me, the resistance is due because the word hardly has any meaning at all unless you precede it with the word 'true' or 'sham', and even then its use is questionable because these terms are subjective (i.e. my 'true' might be your 'sham', and nobody even knows for sure what it was to the player -- which is the case that really matters!)

But I do wonder if there's a cultural split happening here. UK English is such an evolved language that we have no problem accepting ambiguity, sloppy definitions, evolution, difference of opinion. UK English sometimes seems to have far more exceptions than rules.

By contrast, US English is far more orderly and rule-based, both for meanings and pronunciations.

So when people from opposite sides of the pond argue about definitions, the UK contingent tend more to ambiguity and evolution, and the US contingent insist on adherence to rules and definitions. It doesn't mean that either side is right or wrong. It's just a difference of approach.

I've developed this thought a little more on <domdaniel>'s forum, if you're interested.

So I actually have a lot of respect for people who say that the meaning of "sacrifice" is set because a particular authoritative book says so. Or advance a logical argument as you have done. In some cultures those would be compelling arguments.

But I can also respect the argument that the meaning of the word is ambiguous and evolves as usage changes.

I think that is partly the reason that this debate has become so heated - and probably will be heated whenever it comes up. The difference is not so much about the words themselves, it is about how different cultures handle ambiguity versus precision.

Perhaps you're right that it's a cultural thing, although it might also be a 'professional' thing (for lack of a better word). Those of us steeped in fields that depend on precise communication may have less tolerance for slippery definitions.

The extreme case is software development (my profession) where you must communicate precisely with a machine, and there is no room for ambiguity.

I also have a background in mathematics, and there too, great care is taken to formally construct definitions.

From my perspective then, if the practice of making carefully constructed definitions is necessary in cases where clear communication is important, then it must be a Good Thing in general. And yet, I would not want to deprive language of its colorful historical, cultural, and sometimes wacky aspects either.

A very good point, and I fully agree. As well as differences in approach between nations, there are also clear differences between professions.

For example, I work with engineers and politicians. The engineers usually deal in precision terms and rules, where the politicians are more interested in emotions and stuff that can't be measured. I often have to act as a buffer between the two - the engineers work for me, but the politicians are the ones who make the big decisions.

I find that I have to use a different language when I talk to engineers compared to talking to the politicians. The two groups also often take very different ways to tackle a problem.

And don't even get me started on lawyers and media/ marketing experts...

The idea of the sham sacrifice, I think, came from the tendencies of patzers/noobs to label any kind of combination beginning with a material giveaway as a sacrifice. If my combination starts with QxPe7 and ends with me winning a pawn, that doesn't mean I sacrificed a queen. That was just the first, admittedly dramatic, step toward my goal of winning a pawn. But people who haven't reviewed games involving "true" sacrifices just see 15. Qxe7 Nxe7 and think, "OMG he just sacked his queen holy crap!!!" Combinations which begin with a material giveaway and end with a forced mate/win of material cannot possibly be said to be a sacrifice.

That's where all this started! Seems like a long long time ago, in a galaxy far far away ...

A quick recap (and a bit of a summary). I've spotted four different definitions of sacrifice:

1. Any move which temporarily gives up material is a sacrifice.

2. A true sacrifice gives up the material for some considerable time and for an incalculable benefit. A pseudo or sham sacrifice gives up material as part of a calculable line (often leading to a forced mate or material win).

3. There is no such thing as a sacrifice, because with perfect play any position is either a forced win or a forced draw. Admittedly, this is a minority view!

4. There is no clear definition of sacrifice and any of the above could be true.

We've tried several different ways to "prove" that one or other of these is correct. First there's the proof by authority. The definition of true/ pseudo sacrifice is given in Rudolf Speelman's book "The Art of the Sacrifice" and also, BTW, in Wikipedia's entry on chess sacrifices. One point for definition 2. But then again other books use the simple "any deliberate material loss is a sac", so definition 1 takes a point too. Then we started arguing about whether one book was more authoritative than another...

Then there's the dictionary argument. The word "sacrifice" involves giving something up or enduring a hardship, which kind of points to point 2 because a line that leads to a forced mate ain't a hardship. I think that most closely fits your thinking.

Point 3 was argued on a point of logic. Given a perfect super-computer, nothing is incalculable so we cannot base our definition on whether the benefits can be calculated or not.

<yourang> argued on the basis of the need for precision. Definition 1 is simpler to understand and less subjective so it ought to be the one that we choose. After all, whether something is true or sham depends on how much of the follow-up the player saw. This makes for a messy definition. An amateur's true sacrifice could be a computer's or GM's sham sacrifice (because they have seen further).

A number of people argued on the basis of democracy and usage. The most popular definition should prevail. This sort of kills version 3, but is not so clear whether 1, 2 or 4 wins this point.

Language is constantly changing and evolving, so stack up a point for 4.

And finally, a number of people argued on the basis of personal feeling. This tended towards 1, but was not conclusive.

So what's the verdict from all of this? I'd like to posit a fifth version:

"There is no agreed definition of sacrifice. Some believe in a distinction between pseudo and true sacrifices; some don't. Until and unless we have an agreed definition, we each need to make up our own mind about which definition we prefer. And we all need to respect the fact that others may not share our point of view."

Yes it does most closely align with my thinking. I would prefer, however, that we get rid of the term "sacrifice" when talking about pseudo-sacs. Because you're right, there is no hardship in sacking the queen on move 23 when it leads to mate on move 26. It's simply part of the combination.

I like the way Soltis described Tal's approach to sacrifice. "Compensation is in the eye of the beholder. Whereas some players would never consider the position after 18 Nd5 acceptible for white because there are no concrete variations that show white winning, Tal reasoned the other way: There are no concrete variations that show black consolidating." (The game was Tal-Letelier, 1963)

Soltis is touching on, for me, the essence of sacrifice: giving away material for non-material compensation where no forced win is in view.

That's why I quite like "sham sacrifice" or "pseudo sacrifice" - in other words, it only looks like a sacrifice.