Geo-Tagging the Nation’s Homophobia

Hate

A new site takes a look at hateful tweets and where they come from, charting them based on homophobia, racism, and disability. The above map is a graphic representation of the combined tweets using "Dyke", "Fag", "Homo", and "Queer". The reds represent greater intensity.

Looks like western Illinois is a hotbed of homophobia. At least on Twitter.

Comments

  1. Mike8787 says

    This map seems less a map of homophobia concentration and more a population map.

    For the data to be valuable, it would need to chart this information against area population — i.e. how many homophobic tweets in an area per individual, on average.

  2. Joe says

    You also have to take into account that it’s, ya know, Twitter. The playground of bored 13 year olds. Not to say homophobia on it isn’t worthy of noting, it’s just as an Illinoisan, I don’t ever get the sense that I live among a hotbed of homophobia. Quite the opposite.

  3. Rich F. says

    @ Mike8787: Not exactly. If you look at the Boston-NYC-Philadelphia metropolitan axis, it’s at most only light blue (indicating “low hate”), while population density is quite high. There’s almost NO coloration at all in the Los Angeles and San Francisco areas.

    If you click through the link and actually look at the map, you can zoom in and see individual counties. By and large, the red zones DO NOT correlate to major population centers.

  4. says

    That’s not exactly true, Mike8787.

    At the website, you can open the map in a separate window and zoom in. High population centers like Miami, Atlanta, New York, Boston, Chicago, and California’s coastal cities are relatively clear. But if you check out Brownwood, Texas, Sioux City, Iowa, and relatively unpopulated areas of the old Conferacy, not to mention Idaho, you’ll see plenty of red.

  5. Mike says

    Yeah, right. Cuz Illinois, Minnesota & Wisconsin are far more conservative than Idaho, Montana, Wyoming & Mormon Utah, right? Using ‘queer’ doesn’t suggest homophobia. Quoting someone using ‘fag’ doesn’t either. This is a very unlearned, ignorant “study”.

  6. Mike8787 says

    That’s all fine and good. I’m just pointing out that more densely populated states are more colorful, while the entire Mountain time zone is blank — an area that is low in population but likely high, per capita, in homophobic tweeting.

    This also does not account for socio-economic and cultural issues. Tweeting is not a universal activity – certain groups are much more likely to tweet than others.

    I guess the point I’m trying to make is that this map is of very little use. All it shows is “anti-gay tweets happened here.” It provides no lessons about how homophobia in general is concentrated — just how homophobic twitter users are concentrated, which is a useless statistic, in my opinion.

  7. Dastius Krazitauc says

    They should have included Spanish language slurs. I would like to see how the color of Miami changes when you include the word, “maricon”.

  8. greg says

    Huh, the West impressed me today. I’m surprised and heartened. OR all the gays ran away to avoid becoming another Matthew Shepherd and those left are under the delusion that there aren’t any gays here…

  9. Tyler says

    Those of you smart scientists bitching about how the map needs to be adjusted for population and tweets should read the article before posting:

    “Hateful tweets were aggregated to the county level and then normalized by the total number of tweets in each county. This then shows a comparison of places with disproportionately high amounts of a particular hate word relative to all tweeting activity. For example, Orange County, California has the highest absolute number of tweets mentioning many of the slurs, but because of its significant overall Twitter activity, such hateful tweets are less prominent and therefore do not appear as prominently on our map.”

  10. Tyler says

    Those of you smart scientists bitching about how the map needs to be adjusted for population and tweets should read the article before posting:

    “Hateful tweets were aggregated to the county level and then normalized by the total number of tweets in each county. This then shows a comparison of places with disproportionately high amounts of a particular hate word relative to all tweeting activity. For example, Orange County, California has the highest absolute number of tweets mentioning many of the slurs, but because of its significant overall Twitter activity, such hateful tweets are less prominent and therefore do not appear as prominently on our map.”

  11. northalabama says

    ok, i admit, i zoomed in on my town, and relief – no shading, and not much near us, either.

    i guess it’s a good measure of which areas people feel comfortable enough to tweet their discrimination publicly.

    i worry more about those whose discriminate in silence, the ones in the un-shaded areas. you can’t identify them as they are busy working behind the scenes against causes you support. i would almost prefer they were open with their hate. almost.

  12. anon says

    The bad tweets versus normal tweets ratio is not a valid assessment either, as it does not account for the local variation in the use of Twitter. How Twitter gets used and for what purpose would swamp out marginal traffic purposes, thereby greatly affecting the result. A better ratio might be negative versus positive + negative tweets, but determining what a positive tweet looks like is not easy.

    To make the issue clearer, I would imagine in LA, celebrity tweeting about their project might make up 50% of all traffic, but no so in Miami or Atlanta. Thus, gay-neutral traffic in LA might be a lot higher than in those two places simple as a result of how Twitter gets used.

  13. Caliban says

    I saw an article about this yesterday and I have doubts about its validity. Not about the amount of homophobia in our culture but that Twitter can be used to track it in any meaningful way.

    First off only about 13% of the population uses Twitter. And most of those people aren’t going post homophobic or racist slurs even if they are in fact racist or homophobic. So you’re dealing with a very small minority of Twitter users already.

    In fact that it was even possible for a group of students to go through them each “by hand” tells you about the relatively low number of tweets involved. (Facebook comments might give a more representative sample but there are too many of them to go through individually, for example.)

    In that scenario a very small number of people posting & retweeting numerous racist or homophobic slurs could create a “hotspot.”

    So I don’t think this really gives us much meaningful data.

  14. jamal49 says

    Somebody needs to get Nate Silver involved in this. If there are any statistics, they surely are not represented in the map. This is ridiculous. You can’t tell a thing from this graphic representation of this alleged “survey”.

  15. Charlie says

    It’s not very used at the least resolution. It gets interesting when you zoom in.

    I can’t figure out why when you look at the DC area in the greatest resolution there is just a tinge of light blue over Arlington County but whey you zoom out the whole region is red.

  16. ratbastard says

    This study is disingenuous.

    As for why the east, midwest, west Texas, is lit up like a Christmas Tree, that’s because this is where the greatest population densities in the country are located. The west is still by and large barren except for a few urban centers here and there, and along the coast.

    I do find it curious southern California isn’t more lit up. That’s strange. Could it have something to do with the huge Mexican and Hispanic population and their use of Spanish and Mexican/Central American slang?

    ================

    And of course, people especially young people use derogatory words like the N-word all the time. Just because a ‘bad’ word is used doesn’t mean it was a deliberately bigoted tweet.

  17. ratbastard says

    Also interesting: The areas most red in the southeast and parts of the midwest also have huge black populations. This would explain the widespread use of the N-word.

  18. CHRISTOPHER I says

    The article indicates that the context of the tweets used in the study were evaluated by actual humans to ensure that they didn’t count any reclaimed/non-bigoted use of words like “queer” or the n-word.

    It helps to read things before commenting on them.

  19. ratbastard says

    @Christopher,

    You’re correct in my case, I just skimmed it.

    Than the number of tweets was quite small. Again, it’s a disingenuous study. Done to attract attention for the authors.

  20. Paul R says

    Sorry, it still seems more like a map of Twitter use. I don’t know anyone who uses Twitter, and I live in San Francisco—a supposed hotbed of the technology. I also find it hard to believe that DC is incredibly homophobic and the Carolinas and Florida are not, having spent plenty of time in all four places. Or that the left half of the country is essentially free of homophobia.

  21. says

    Like others have said, I’m not sure that this map is extremely reflective of levels of homophobia considering differences in population density, percentage of the population using Twitter, etc. If anyone remembers the map of number of Facebook profiles pictures changed to support equality, I think that that map was more representative of overall attitudes than this one…

  22. Bill says

    @Paul R: the map does not show that “DC is incredibly homophobic,” but you have to go to the web site and zoom in to tell.

    @ratbastard: it is not a “disingenuous study” but rather some data displayed in the form of a map. It’s useful: the large regional variations suggest that controlling for that is important in any real study of homophobia.

    At least in the physical sciences, a lot of effort goes into experiments to make sure the results aren’t being skewed by various errors, and the vast majority of that does not appear in publications – if it did, the publications would be many times longer.

  23. EchtKultig says

    “So many people on this thread extrapolating from their personal experience. Your personal experience is worthless, sorry.

    Posted by: Mike B. | May 13, 2013 3:26:48 PM

    Yep. And apparently no one is learning critical reasoning in school anymore. You don’t need to bias the map for population density because twitter users are (obviously) already a self-selecting group. The map DOES show exactly what one would expect.

  24. EchtKultig says

    “I’m just pointing out that more densely populated states are more colorful, while the entire Mountain time zone is blank — an area that is low in population but likely high, per capita, in homophobic tweeting.”

    Well, perhaps Towleroad shouldn’t have titied the post “geo tagging hate”. The study only purports to track hate on twitter. So, of course, the middle of Utah might have only homophobic residents – anyone should know that anyhow – but perhaps there aren’t ANY homophobes who are tweeting in those areas, because they aren’t tweeting at all. That’s why the area shows up as null on the map.

  25. Tatts says

    On a national scale, this is a very, very, very small sampling when aggregated at the county level (there are over 3,100 counties in the US). In some counties, the entire “hate” may be coming from one individual.
    Plus, it only is triggered by a few of the worst, most obvious key bad words. Racists who talk about “Obama supporters”, etc., instead of using the N-word fly under the radar of this “study”.
    The study authors acknowledge that this is only their first iteration of the study, so don’t read anything into it. It will be a long time before they finish tweaking their methodology.

  26. Bill says

    @Paul R: What “findings?” It is a graphical representation of homophobic “tweets” per county normalized to the total number of “tweets” from that county. That’s all it is. The people who provided the map didn’t draw any conclusions from it. They just showed the map (an interactive one where you can ‘zoom’ in).

    Some of the questions the map raises are the following:

    1. Are twitter users in a county representative of that county?

    2. In the areas high on the ‘hate’ scale, are the hateful tweets from people who know each other (e.g., all in a “gays are going to Hell” church) or is there some local loudmouth they all listen to, or is there some other association between them?

    3. Do hateful individuals put out more tweets than more reasonable individuals, thereby skewing the data?

    And that is just for starters. The good thing about the map, even in such a preliminary form, is that it suggests a series of things that might be worth investigating in detail.

Leave A Reply