Sorted by New


2014 Less Wrong Census/Survey

are you saying that because the definition of a heap is vague, there are multiple feasible definitions of a heap, so some people (using one definition) would call a candidate heap a heap and other people (using another definition) would say it isn't a heap?

Not that they are necessarily using multiple definitions, but the common definitions themselves do not specify an exact range in quantity in which a cluster of things could be considered a heap. Two people could disagree about whether something is a heap despite using the same vague definition of "heap". They may be comparing the candidate heap relative to things that they have experienced being called heaps in the past. I suppose you could still treat this as being a difference in their personal definitions of "heap". However, I don't think that if pressed to define "heap", that people would be likely to state an explicit quantity range. They would most likely give vague qualitative definitions. The same person may even use inconsistent definitions at different times or forget to include certain aspects that they would consider to be important defining characteristics. People don't normally think in terms of definitions when classifying things. They usually just classify based on what feels correct, and definitions are after-the-fact attempted explanations of their classifications.

Almost all terms introduce a simplification, not just the vague ones.

Doesn't the introduction of a simplification itself give a term some vagueness, because then you don't know the details of the relevant characteristic, just a qualitative judgment? For "heaps", the simplification is one that keeps the exact quantity obscured while providing a qualitative description instead. Even if you had a different quantity term that, unlike "heap", didn't have fuzzy boundaries, it could still be considered vague in a different sense if multiple quantities could fulfill its definition (assuming an exact quantity really did exist in reality). It would certainly at least be considered somewhat ambiguous. For example, the category "integers" has seemingly clear boundaries, but calling an unknown number an integer is still vague if it doesn't express all the relevant information.

vague terms have a wider array of meanings than non-vague terms

Yes, terms can be vague in that way too.

2014 Less Wrong Census/Survey

I could still call certain things heaps even if I had a sharp definition of a heap.

Sorry, what I meant was, "Heaps as they are currently defined can be said to exist because of vagueness in the definition of what precisely makes something a heap." They can also be said to not exist because of vagueness in the definition of what precisely makes something a heap. However, it is extremely difficult if not impossible for any definition to have absolutely no ambiguity at all, and even if that was possible, language can still be considered a social construct in the sense that linguistic terms are constructed socially. Vague terms have an extra layer of social construction though, because rather than just giving a term to a phenomenon, they also introduce a simplification. This makes the constructed aspect more obvious. I guess I didn't word my above quote very well, since I didn't mention that the social construction of all language itself is also enough to make "heaps" a social construct.

2014 Less Wrong Census/Survey

I'm just using it to mean things that are constructed socially.


2014 Less Wrong Census/Survey

"Social construct" implies an arbitrary choice -- our society decided to split humanity into races this way, but another society might do it in an entirely different way and all such ways are equally valid, which is to say, there are no underlying "real" differences.

Supposing someone wanted to split humanity into arbitrary races based on actual genetics (which is not how the concept of race originally started because genetics wasn't known at the time), it would make sense for most races to be African, since Africa has far more human genetic diversity than all the other continents combined do. The reason races are delineated the way they are now is due to social reasons. (It could possibly make sense when you consider the phenotype though, but due to the outgroup homogeneity bias, I have some doubts.)

Still, regardless of where you set the boundaries between races, there will be average biological differences between them (provided you don't do something biologically ridiculous like classifying whites and Asians as the same race but then classifying their half-white/half-Asian children as a different race).

2014 Less Wrong Census/Survey

Part of the reason is that if you restrict to the population of the United States they are (more-or-less) a separate genetic cluster

Well, I wasn't restricting to the population of the United States. Anyway, race is still a socially constructed identity. This is apparent with mixed-race people who often identify with one race more than another based on how they were raised, how they look, how other people identify them, and whether they act more like a stereotypical member of one of their races than another. The race they identify most with might not be the one that makes up the largest proportion in their ancestry.

Only because anyone who dares to point out the obvious truth that it isn't gets called a "sexist transphobe" and unfit for polite society.

My understanding is that gender is specifically used to refer to the socially constructed identities. Biological sex differences get lumped under sex rather than gender, which is why people can believe in the social construct of gender while also believing that biology contributes in some degree to stereotypical gender roles. I'm not an expert on gender though, so I should probably leave it to someone else to debate you on this point.

2014 Less Wrong Census/Survey

From what people have said, it seems that after the survey was posted a new question was added about our favorite LW post. Were there any others?

(Posted as a top-level comment at the request of TobyBartels)

2014 Less Wrong Census/Survey

From what people have said, it seems that after the survey was posted a new question was added about our favorite LW post. Were there any others?

2014 Less Wrong Census/Survey

I never said race wasn't a useful concept. I specifically said in my earlier post: .

I'm not saying that "heap" and "race" are not useful terms. They do correlate with actual differences,

I think my initial post that started this discussion may have been a source of misunderstanding. When I called race a social construct, I wasn't trying to say that race is a useless concept, but instead indicate that it could be useful as a cultural/identity concept. Initially when I talked about "mixed race" and "Hispanic" not technically being races, I was defining race according to the mainstream definition that treats race as a genetically distinct group of people, since that is my default. However, during the part where I talked about how Hispanics are often treated as if they were a race, I was undergoing a shift toward thinking about race as a cultural identity regardless of genetics, which then led me to the statement that race is a social construct. I meant it in a similar way to what people mean when they say that gender is a social construct. When people say that, they're not implying that gender is a useless concept, but that it is a personal subjective choice of identity. Significantly, I then spent the rest of my post talking about race as a personal choice of identity.

The idea that gender is a social construct is a pretty uncontroversial one, as far as I can tell. People seem to be somewhat less likely to say the same thing about race though, probably because "race" as a cultural term doesn't have a satisfactory parallel term to refer to biology the way "gender" has "sex". It didn't matter for me in practice though. I thought of race as a social construct regardless of whether it was approached from a biological or cultural perspective, which is why I didn't feel a need to distinguish between the two in my statement. However, subsequent comments drawing attention to its biological validity (e.g. would doctors agree?) pushed me to address my point underlying my passive implication that the biological aspect is also a social construct, which then skews the discussion in a way that buries much of my original meaning. The social construction of race as a biological concept is not itself adequate to explain why I would support including non-genetic race answers to a race question, but the social construction of race as a subjective personal identity is.

Earlier I was wondering why my comments were getting downvoted. What could possibly be so controversial about the idea that human genetic variation is a continuum, or that linguistic terms are socially constructed? Now I can see that if these are interpreted as if they are supposed to be arguments in support of including non-genetic answers to a race question or a lack of average differences between races, they might seem like bad arguments, but I wasn’t intending them to support those premises, and I didn’t think that people would think I was intending them to.

Load More