"So here’s my question: With the “ethnicity” question, is this entrepreneur courageously addressing the proposition that we’re different according to our ancestry, and propelling us toward a post-racial future? Or is he pretending to be scientific as a marketing gimmick, while actually enforcing false, outdated and possibly dangerous ideas about race?"
No. He found out that if he asked questions about ethnicity to the customer he actually could size them better. Everything else in this article is in your head.
This is actually a well written and researched and interesting article about race as a social construct and as a physical reality. Perhaps you shouldn't just dismiss it because you (apparently) can't be bothered to think about these things. Also it's not exactly uncommon (or reprehensible) to hang an article like this off of a pithy and slightly overstated sound bite.
First, historically, race was only firmly established as a legacy of the colonial era. For example, in the colonial era, how was it decided in Louisiana if one man was black or not? The answer is, less than 1/32 of his ancestry must be of "black blood"[0]. This sort of thinking and arbitrary ruling is how race was established as a social construct.
Science at the time, based on the physiology of skull shapes and other body features, was shaped in order to draw a line between the "superior" race and the "inferior" race; this is a belief we sometimes still carry. In modern science we know that there is enough variation in people that it is basically impossible to concretely associate certain genes with certain races[1]. So in fact, race IS a social construct; race is the generalizations we unfairly make.
[As a bonus, any two humans average 99.5% the same DNA]
That doesn't refute the parent's point. There is more to race than some jackass using it as a means to discriminate. It is not just a "social construct".
I'm saying race is based on historical and social circumstances. This notion is further supported because races are scientifically (genetically) difficult to distinguish. Hence "race is a social construct" is in fact based in science, directly arguing against the parent comment.
This is especially silly and misleading. Race is considered a "social construct" under the theory of Social Constructionism. For people who don't subscribe to the theory of Social constructionism, race is not a "social construct". Social constructionism itself is not based on biological and evolutionary fact. It is a philosophical approach to the world that conveniently disregards biology, genetics, behavioral ecology, and enormous scale (n=millions) of statistical studies. When you remove Social Constructionism from your thesis, your entire argument falls apart.
Most people do not believe in discredited theories like social constructionism. The other posters in this thread are asking you to rework your argument without appeals to Social Constructionism, and you show you are unable to do so.
Races are not difficult to distinguish, in fact determining self-identified race from genetic data is trivial, and genetic data is clustered in ways that reflect the common usage of racial categories. For example see http://www.gettinggeneticsdone.com/2011/10/new-dimension-to-...
>First, historically, race was only firmly established as a legacy of the colonial era.
I'm not sure what you're trying to say here. Are you saying a Roman in Africa didn't notice the guys around him had black skin?
>For example, in the colonial era, how was it decided in Louisiana if one man was black or not? The answer is, less than 1/32 of his ancestry must be of "black blood"[0]. This sort of thinking and arbitrary ruling is how race was established as a social construct.
This is non-responsive. You're talking about the legal response to race, which is irrelevant to the question of whether race actually exists.
In fact, you've written nothing that addresses the point. When a white guy and a black guy are walking down the street, do you have trouble when someone asks you to point out the white guy?
>I'm not sure what you're trying to say here. Are you saying a Roman in Africa didn't notice the guys around him had black skin?
"Romans in Africa vs. black skin" is actually a different point. Differences existed between people, but the notion of "race" did not. I'm trying to explain what the history of the notion of "race" was, because we take it for granted now. If you would read the history of race, especially the colonial time around the 17th to 19th century[0], you can see that that is when the notion of "race" as we know it really started to develop as a result of anthropological studies and colonial goals (but NOT genetic, modern scientific studies).
>For example, in the colonial era, how was it decided in Louisiana if one man was black or not? The answer is, less than 1/32 of his ancestry must be of "black blood"[0]. This sort of thinking and arbitrary ruling is how race was established as a social construct.
This is an example of how arbitrary and socially constructed race is.
Not sure if you read my second paragraph, but I wrote this: "In modern science we know that there is enough variation in people that it is basically impossible to concretely associate certain genes with certain races[1]."
You seemed to be convinced that "race is a social construct" has no scientific basis, but my argument is that there is no strong basis for race in science. The only alternative choice then seems to be social.
The term "race" was most likely coined by Immanuel Kant, in German as rassen, to name his view that there were actually different kinds of people. The more common view before Kant was that environment shaped people.
It's not really true that race is a "socially construct". It's a specific belief about the nature of human beings, attributable to a specific person.
As a fun side note, being "racist" means believing that race is a real thing. It has nothing to do with what one thinks of different races, how one feels about them, or whether or not one feels differently about them at all. Just be living that "race" names something true about humans makes a person a racist.
My biggest problem with the article was just a few paragraphs into it, the author claims that the shirt company early on discovered "They seemed to fit worse, in a predictable way, on people who weren’t Caucasian." And then the author spends the next 80% of the article discussing whether race is an illusion and whether data about race being different is typically wrong.
BUT the author started off by saying that basically by asking this question, the company can make the shirts fit better.
Isn't that the answer then? We are all different, without perfect custom tailoring, you can't really mass market to fit everyone. Not everyone who is 5'11 will wear the same size shirt. Nor will everyone who is "black."
The author takes great pains to point out that we are a range. But the author also pointed out at some point you say "that person is 'Asian'"
But if this company figured out a way to make the shirts fit better simply by asking for the ethnicity, then what is the problem?
Did you read the whole thing? The conclusion of the article seems to be the same as that of your parent poster, even if the article's author has gotten there in a more rigorous manner.
Reading through, it feels like the "other side" is guilty of the same bias you accuse your parent -- the so-called experts referenced in the article who are on the side that race doesn't exist... they came off as ignoring actual real data (and when presented real data, hand-waving it away) in favor of what they want to believe as the ideal. And I'd agree with that ideal, that race shouldn't exist as a social construct -- but! -- that doesn't preclude race as being a predictor of certain body shapes and types.
Saying "white people tend to have longer torsos and shorter legs, while black people tend to have shorter torsos and longer legs" isn't racist. Even if it's wrong, why must it be racist? I mean... what possible racially-biased motivation could you have to compare torso vs. leg length in different ways? There are far easier, more obvious ways to be racist, if that's your thing. But, regardless, the data here seems to suggest that there actually is a difference, so... well, there we are.
It looks like in his studies the results have been that ethnicity [in combination with people's guestimates?] is a better predictor of a person's shirt size than people solely providing their own measurements. That's definitely an improvement. It's possible he could use Kinect or the like to get even better information for creating better fitting shirts.
Still, if this current method is more effective in predicting shirt sizes, I really see no reason not to use it.
Would people have an issue with integrating more diverse ethnicity studies into pharmacological studies? Probably not. So I don't see why it would be different here.
One of the references cited in this article is a line of shoes released by Nike designed to fit Native Americans better. They measured 200+ native feet and determined that they were wider toward the toe than typical shoes.
While I don't culturally identify as native, there's no doubt that I am 1/16th Cherokee. As it turns out, I have what I call "duck feet". They are very wide up front. I didn't know about Nike's foray into native shoes, but I wish I had and that they still made them.
Rather than being offended, I would be excited to have some shoes that actually fit well. I usually end up with shoes that too big in other dimensions or toe-crampers.
Try New Balance, they too have some interesting dimensions along this line, but I'm not sure on the origin. I would be curious what tribes Nike selected for their study and if the data is available.
I find it depressing that the sociologists are saying that race is (only) a social construct. I mean, sure, there's an element of truth to what they're saying, but there's also a truth that they're denying. We build these social constructs because there's an external reality that they... well, let's say that they oversimplify. But the external reality is there, and it isn't going away.
The external reality is that there are real, physical differences between people that correlate at least somewhat with what we perceive as "race". That's reality, whether sociologists choose to acknowledge it or to ignore it.
So the problem with categorizing people into races is that there are many axes along which people can be divided into buckets, and any particular race grouping only has meaning in a particular place and time and will have totally arbitrary boundaries, which are as much cultural as they are genetic and physical. As long as people understand that, it's okay to talk about races tending to have certain qualities. Because culturally we've divided people based on those qualities.
But it's really important to understand that people of different races aren't different kinds of people. Any more than people who have blonde or brunette hair are different kinds of people.
>The external reality is that there are real, physical differences between people that correlate at least somewhat with what we perceive as "race".
This doesn't seem to contradict what they are saying here though.
From the article:
>Here’s what they mean: if you were to travel across the Eurasian continent from Portugal to Japan, say, there would be no river, forest or discrete boundary where people suddenly started looking “Asian” (or “European” if you traveled from east to west). Instead, changes in physical appearance would occur so gradually and imperceptibly that you probably wouldn’t notice. You’d only be aware that, once in Japan, people definitely looked different than in Portugal.
As far as I can tell, saying that race is a social construct just the age-old insight that the way we slice up the world into categories tends to be somewhat arbitrary. Race is a social construct just like species is a social construct. Why are wolves and dogs different species? You can come up with the most precise definition of species you want (which is difficult) to explain "why", but any possible definition you come up with is still somewhat arbitrary. You could have chosen a different definition which would make them the same species, or you could have split the (wolf + dog) category into three categories. This doesn't require anybody to ignore the differences between wolves and dogs. It just suggests that maybe we shouldn't get too attached to thinking of a husky as a dog, maybe it should be thought of as some in between species?
I suspect the problem could be eliminated by taking a few more measurements.
Whatever is systematically wrong that correlates with race is a form of error. You know that it's wrong because it is measurable. If it is measurable you can just turn it into a measurement which compensates for the error.
How about giving users a choice: workflow A with fewer measurements plus race question, or workflow B with a couple more measurements and no race question.
I suspect you have to get a circumference around the shoulders, the chest under the arms, the waist around the navel and around the hips-butt. Also, the circumference of the upper arm and neck circumference. If you have all these, I don't see how you can possibly mess up a T shirt fit.
Ignoring the racial issue for a moment, I don't understand how they ended up with the questions at all.
> Naysayers told him that when customers input their measurements, they often made mistakes — the idea wouldn’t scale.
> Asking about waist size was insufficient, for example, because it gave no indication of the size of one’s midsection. So Skerritt added a question about how far one’s belly protruded. Other questions were too confusing, like one about how T-shirts fit around your chest and shoulders.
1. There are dozens of fitted clothes manufacturers. More every year. It doesn't look like they fail, or not scale. I see almost only positive reactions about them on twitter.
2. Of course trivial measurements will not be enough. The article gives the example of midsection. But that's not other services ask for. Tailorstore asks for either common t-shirt size, 2-3 measurements, or if you actually want a fitted shirts, you provide: neck, chest, waist, hip, seat, shirt length, shoulder width, arm length, wrist. No need to ask about hanging bellies - just ask for measurements.
When they ask extra questions they seem to derive some ratios between basic sizes. Maybe it works, maybe not... but did they actually find out that taking own measurement doesn't work? I ordered some tailored shirts online, gave all my measurements and couldn't be happier with the result. Are the extra questions actually worth it? Maybe they did verify that, but it's not in the article, and I find that really annoying - here's the solution to what people told us will be the problem, we're not going to tell you if we verified that it is a problem.
You're probably better with a tape measure than I am. I'd have trouble putting a tape measure on my shoulders and positioning it in the right place — my shoulders are rounded, not angular. I'd probably screw up most of the other measurements as well, and I already have a tape measure. What about the ≈99.9%* of guys who don't? This sounds like a service for them.
* I made this number up just now, may be off by an order of magnitude or two
I used the metal, rolled up one, mostly used for DIY stuff. It worked :) (definitely wasn't precise though) You can also use one in any shop with shirts. I don't think it's that big deal really.
I wonder if this kind of reporting is sponsored somehow by a competitor.
---
He was essentially saying, Prove that it exists. And if you can’t or won’t prove it, then stop talking about it. Because without proof, the concept of race, as it pertains to variation in the human family, is too dangerous.
Let’s open a conversation about this. I’m curious to hear people’s lived experiences in this arena.
---
Well the author already opened the conversation. Next up is probably a storm of Twitter responses casting this comapany as "racist".
It's interesting that despite trying to be highly sympathetic, the article still paints Goodman in a bad light, in my opinion. All Goodman can say is that race is a continuum, not a finite set. But so what? Self defined race is still highly correlated with genes, and hence phenotype (in this case, body dimensions), so why is self-defined race not a useful category? Anyone with any knowledge of statistics, probability or machine learning can see through Goodman's platitudes. Maybe left-leaning anthropologists have other reasons for disliking these uses of the concept of race, but they need to be more intellectually honest.
Interesting that it was the white folks that got the good fit, out of the box. Kind of makes the race thing seem like a hacked on solution to a fundamentally biased algorithm.
> Interesting that it was the white folks that got the good fit, out of the box.
Assuming their market is primarily in the US, then 72% of their potential customers are white, 12% are black, 5% asian, and the rest something else [1].
That's for the US as a whole. I'd expect the subset of the US population that will buy online custom dress shirts tends toward the more affluent end of the population, where the skew is even more heavily toward white.
So the default option is the one that is best for the majority (by a large margin) of their customers. I don't see what is interesting about that.
... and a similar experiment in predominantly Han Chinese China likely results in the same thing biased towards Han Chinese, In Thailand towards Thais, and in Cambodia towards Khmer people.
My GF lived in China half a decade, and she had trouble with tailored clothes, because they could never accept her southern-European proportions.
The idea that the most common size should be the most common group is just business 101 (build for your customers).
Instead of asking for race/ethnicity, just say that the shirts are available in fit variations A, B, C, etc., and show photos of several models of all races each modeling the variation that fits that model best.
Leave it to the customer to notice that the majority of white models are wearing "A", the majority of black models are wearing "B", and so on and guess that if they don't know otherwise they have the best chance of getting a good fit by going with the variation that is best for the majority of models that look like them.
This still lets customers have the best shot at getting a great fit, and furthermore if it turns out wrong for a particular customer it is easier for them to realize they need to try another variation next time.
If I ordered a shirt and told the company I'm white and it did not fit well, it would not occur to me to go back and tell them I'm black on my next order, because I probably would not even be connecting race with fit.
With an A, B, ..., system, on the other hand, if my "A" shirt didn't fit right it would immediately occur to me that maybe the "B" shirt would be better.
"Instead of asking for race/ethnicity, just say that the shirts are available in fit variations A, B, C, etc., and show photos of several models of all races each modeling the variation that fits that model best."
That assumes people will accurately self-assess and honestly report their body types. And...you'd be very surprised at how few people can do that.
In fact, "vanity sizing" is quite common in the fashion business. This is a term meaning a label that says one size, often smaller than the actual measurements of the garment, in order to make the buyer feel better about him/herself. (For instance, ever buy a pair of jeans in size 34? Chances are they're really a 36. The clothing company figures you'd refuse to admit to yourself that you're actually a 36. So it plays to your vanity, rather than risking your business by telling you something you don't want to hear.)
On the whole, your idea seems better than "race" or "ethnicity" as category. But the weird benefit of "race" or "ethnicity" is that it doesn't force the buyer to take a conscious self-assessment of his body type.
Perhaps another way to go about this would be to ask a series of questions. "Do you ever have the following problems buying shirts off the rack? [Problem 1,Problem 2,Problem 3,...,Problem N]." These would come with visualizations. Example: "The sleeves are usually too short."
I worked in men’s clothing for quite a while, and I have an anecdote to report about customers self-assessing their size and cut. If you sell people what they ask for, they will hate you for it.
In my day, off-the-rack shirts came in a collar size, an arm length, and there were three cuts: tailored or “european,” standard, and “full,” for the pear-shaped gentleman. People sized themselves at least an inch too small on the collar, and would woefully under-report their arm length too. The problem, of course, was that as you gain weight, your neck grows in girth and you need a larger collar.
What about the arm length? Well, that’s measured from the nape of the neck to the wrist. As you gain weight, you gain fat and/or muscle around your shoulders and upper arm, and that can increase the measured length of the arm, even though your skeleton remains the same. Customers would ask for a shirt in the size they wanted their neck to be, and if you sold it to them, they’s squeeze into it, there’s be an unsightly bulge, and they would decide that your shirts were unflattering.
One of the tricks in those days was to report that a particular shirt’s measurement was more accurate in metric, so if the customer didn’t mind, could I take a metric measurement... And I would sell them the correctly sized shirt. Although it was larger than they requested, it would look better because it fit properly, and they would see a thinner, fitter person in the mirror.
Needless to say, the exact same thing was true in the cut of the shirt. The endomorphs would ask for a standard, and the standards would ask for the tailored cut. And then they’d hate the shirt.
The best thing was to report that this particular manufacturer was off on their sizing, and just hand them the short that you could eyeball was right for them. They’d be back for more.
So where were we? Oh yes, you absolutely cannot rely on people self-selecting their body type. The best thing is to measure.
> That assumes people will accurately self-assess and honestly report their body types. And...you'd be very surprised at how few people can do that.
Skerritt actually addresses this, though near the end of the article:
"Respondents only need to answer the question to the best of their abilities. Because people tend to make the same mistakes, or tell the same lies, or have the same misconceptions about themselves — because none of us are as original as we like to think — they invariably produced similar responses, yielding the patterns necessary to fit your shirt."
So it seems it doesn't matter. Even when people lie, they tend to all lie in the same ways, and you can still get a proper fit based on the patterns in the lies.
> That assumes people will accurately self-assess and honestly report their body types. And...you'd be very surprised at how few people can do that.
I don't think that matters in this case. Even if I am way off in my self-assessed body type, and so look for a ripped Hercules of a model when deciding which fit variation to order instead of the Homer Simpson model that I should be looking for...I'm likely to pick a ripped Hercules that matches my race. So, if the fit variation options are just for choosing the value of that parameter that the race question currently answers, it should work the same for most people.
No. He found out that if he asked questions about ethnicity to the customer he actually could size them better. Everything else in this article is in your head.