What’s up with calling a woman “a female”? A look at the parts of speech of “male” and “female” on Twitter .

This is something I’ve written about before, but I’ve recently had several discussions with people who say they don’t find it odd to refer to a women as a female. Personally, I don’t like being called “a female” becuase its a term I to associate strongly with talking about animals. (Plus, it makes you sound like a Ferengi.)  I would also protest men being called males, for the same reason, but my intuition is that that doesn’t happen as often. I’m willing to admit that my intuition may be wrong in this case, though, so I’ve decided to take a more data-driven approach. I had two main questions:

  • Do “male” and “female” get used as nouns at different rates?
  • Does one of these terms get used more often?

Data collection

I used the Twitter public API to collect two thousand English tweets, one thousand each containing the exact string “a male” and “a female”. I looked for these strings to help get as many tweets as possible with “male” or “female” used as a noun. “A” is what linguist call a determiner, and a determiner has to have a noun after it. It doesn’t have to be the very next word, though; you can get an adjective first, like so:

  • A female mathematician proved the theorm.
  • A female proved the theorm.

So this will let me directly compare these words in a situation where we should only be able to see a limited number of possible parts of speech & see if they differ from each other. Rather than tagging two thousand tweets by hand, I used a Twitter specific part-of-speech tagger to tag each set of tweets.

A part of speech tagger is a tool that guesses the part of speech of every word in a text. So if you tag a sentence like “Apples are tasty”, you should get back that “apples” is a plural noun, “are” is a verb and “tasty” is an adjective. You can try one out for yourself on-line here.

Parts of Speech

In line with my predictions, every instance of “male” or “female” was tagged as either a noun, an adjective or a hashtag. (I went through and looked at the hashtags and they were all porn bots. #gross #hazardsOfTwitterData)

However, not every noun was tagged as the same type of noun. I saw three types of tags in my data: NN (regular old noun), NNS (plural noun) and, unexpectedly, NNP (proper noun, singular). (If you’re confused by the weird upper case abbreviations, they’re the tags used in the Penn Treebank, and you can see the full list here.) In case it’s been a while since you studied parts of speech, proper nouns are things like personal or place names. The stuff that tend to get capitalized in English. The examples from the Penn Treebank documentation include “Motown”, “Venneboerger”,  and “Czestochwa”. I wouldn’t consider either “female” or “male” a name, so it’s super weird that they’re getting tagged as proper nouns. What’s even weirder? It’s pretty much only “male” that’s getting tagged as a proper noun, as you can see below:


Number of times each word tagged as each part of speech by the GATE Twitter part-of-speech tagger. NNS is a plural noun, NNP a proper noun, NN a noun and JJ an adjective.

The differences in tagged POS between “male” and “female” was super robust(X2(6, N = 2033) = 1019.2, p <.01.). So what’s happening here?  My first thought was that it might be that, for some reason, “male” is getting capitalized more often and that was confusing the tagger. But when I looked into, there wasn’t a strong difference between the capitalization of “male” and “female”: both were capitalized about 3% of the time. 

My second thought was that it was a weirdness showing up becuase I used a tagger designed for Twitter data. Twitter is notoriously “messy” (in the sense that it can be hard for computers to deal with) so it wouldn’t be surprising if tagging “male” as a proper noun is the result of the tagger being trained on Twitter data. So, to check that, I re-tagged the same data using the Stanford POS tagger. And, sure enough, the weird thing where “male” is overwhelming tagged as a proper noun disappeared.


Number of times each word tagged as each part of speech by the Stanford POS tagger. NNS is a plural noun, NNP a proper noun, NN a noun, JJ an adjective and FW a “foreign word”.

So it looks like “male” being tagged as a proper noun is an artifact of the tagger being trained on Twitter data, and once we use a tagger trained on a different set of texts (in this case the Wall Street Journal) there wasn’t a strong difference in what POS “male” and “female” were tagged as.

Rate of Use

That said, there was a strong difference between “a female” and “a male”: how often they get used. In order to get one thousand tweets with the exact string “a female”, Twitter had to go back an hour and thirty-four minutes. In order to get a thousand tweets with “a male”, however, Twitter had to go back two hours and fifty eight minutes. Based on this sample, “a female” gets said almost twice as often as “a male”.

So what’s the deal?

  • Do “male” and “female” get used as nouns at different rates?  It depends on what tagger you use! In all seriousness, though, I’m not prepared to claim this based on the dataset I’ve collected.
  • Does one of these terms get used more often? Yes! Based on my sample, Twitter users use “a female” about twice as often as “a male”.

I think the greater rate of use of “a female” that points to the possibility of an interesting underlying difference in how “male” and “female” are used, one that calls for a closer qualitative analysis. Does one term get used to describe animals more often than the other? What sort of topics are people talking about when they say “a male” and “a female”? These questions, however, will have to wait for the next blog post!

In the meantime, I’m interested in getting more opinions on this. How do you feel about using “a male” and “a female” as nouns to talk about humans? Do they sound OK or strike you as odd?

My code and is available on my GitHub.


What’s the difference between & and +?

So if you’re like me, you sometimes take notes on the computer and end up using some shortcuts so you can keep up with the speed of whoever’s talking. One of the short cuts I use a lot is replacing the word “and” with punctuation. When I’m handwriting things I only ever use “+” (becuase I can’t reliably write an ampersand), but in typing I use both “+” and “&”. And I realized recently, after going back to change which one I used, that I had the intuition that they should be used for different things.


I don’t use Ampersands when I’m handwriting things becuase they’re hard to write.

Like sometimes happens with linguistic intuitions, though, I didn’t really have a solid idea of how they were different, just that they were. Fortunately, I had a ready-made way to figure it out. Since I use both symbols on Twitter quite a bit, all I had to do was grab tweets of mine that used either + or & and figure out what the difference was.

I got 450 tweets from between October 7th and November 11th of this year from my own account (@rctatman). I used either & or + in 83 of them, or roughly 18%. This number is a little bit inflated because I was livetweeting a lot of conference talks in that time period, and if a talk has two authors I start every livetweet from that talk with “AuthorName1 & AuthorName2:”. 43 tweets use & in this way. If we get rid of those, only around 8% of my tweets contain either + or &. They’re still a lot more common in my tweets than in writing in other genres, though, so it’s still a good amount of data.

So what do I use + for? See for yourself! Below are all the things I conjoined with + in my Twitter dataset. (Spelling errors intact. I’m dyslexic, so if I don’t carefully edit text—and even sometimes when I do, to my eternal chagrin—I tend to have a lot of spelling errors. Also, a lot of these tweets are from EMNLP so there’s quite a bit of jargon.)

  • time + space
  • confusable Iberian language + English
  • Data + code
  • easy + nice
  • entity linking + entity clustering
  • group + individual
  • handy-dandy worksheet + tips
  • Jim + Brenda, Finn + Jake
  • Language + action
  • linguistic rules + statio-temporal clustering
  • poster + long paper
  • Ratings + text
  • static + default methods
  • syntax thing + cattle
  • the cooperative principle + Gricean maxims
  • Title + first author
  • to simplify manipulation + preserve struture

If you’ve had some syntactic training, it might jump out to you that most of these things have the same syntactic structure: they’re noun phrases! There are just a couple of exception. The first is “static + default methods”, where the things that are being conjoined are actually adjectives modifying a single noun. The other is “to simplify manipulation + preserve struture”. I’m going to remain agnostic about where in the verb phrase that coordination is taking place, though, so I don’t get into any syntax arguments ;). That said, this is a fairly robust pattern! Remember that I haven’t been taught any rules about what I “should” do, so this is just an emergent pattern.

Ok, so what about &? Like I said, my number one use is for conjunction of names. This probably comes from my academic writing training. Most of the papers I read that use author names for in-line citations use an & between them. But I do also use it in the main body of tweets. My use of & is a little bit harder to characterize, so I’m going to go through and tell you about each type of thing.

First, I use it to conjoin user names with the @ tag. This makes sense, since I have a strong tendency to use & with names:

  • @uwengineering & @uwnlp
  • @amazon @baidu @Grammarly & @google

In some cases, I do use it in the same way as I do +, for conjoining noun phrases:

  • Q&A
  • the entities & relations
  • these features & our corpus
  • LSTM & attention models
  • apples & concrete
  • context & content

But I also use it for comparatives:

  • Better suited for weak (bag-level) labels & interpretable and flexible
  • easier & faster

And, perhaps more interestingly, for really high-level conjugation, like at the level of the sentence or entire verb phrase (again, I’m not going to make ANY claims about what happens in and around verbs—you’ll need to talk to a syntactician for that!).

  • Classified as + or – & then compared to polls
  • in 30% of games the group performance was below average & in 17% group was worse than worst individual
  • math word problems are boring & kids learn better if they’re interested in the theme of the problem
  • our system is the first temporal tagger designed for social media data & it doesn’t require hand tagging
  • use a small labeled corpus w/ small lexicon & choose words with high prob. of 1 label

And, finally, it gets used in sort of miscellaneous places, like hashtags and between URLs.

So & gets used in a lot more places than + does. I think that this is probably because, on some subconscious level I consider & to be the default (or, in linguistics terms, “unmarked“). This might be related to how I’m processing these symbols when I read them. I’m one of those people who hears an internal voice when reading/writing, so I tend to have canonical vocalizations of most typed symbols. I read @ as “at”, for example, and emoticons as a prosodic beat with some sort of emotive sound. Like I read the snorting emoji as the sound of someone snorting. For & and +, I read & as “and” and + as “plus”. I also use “plus” as a conjunction fairly often in speech, as do many of my friends, so it’s possible that it may pattern with my use in speech (I don’t have any data for that, though!). But I don’t say “plus” nearly as often as I say “and”. “And” is definitely the default and I guess that, by extension, & is as well.

Another thing that might possibly be at play here is ease of entering these symbols. While I’m on my phone they’re pretty much equally easy to type, on a full keyboard + is slightly easier, since I don’t have to reach as far from the shift key. But if that were the only factor my default would be +, so I’m fairly comfortable claiming that the fact that I use & for more types of conjunction is based on the influence of speech.

A BIG caveat before I wrap up—this is a bespoke analysis. It may hold for me, but I don’t claim that it’s the norm of any of my language communities. I’d need a lot more data for that! That said, I think it’s really neat that I’ve unconsciously fallen into a really regular pattern of use for two punctuation symbols that are basically interchangeable. It’s a great little example of the human tendency to unconsciously tidy up language.

Do you tweet the way you speak?

So one of my side projects is looking at what people are doing when they choose to spell something differently–what sort of knowledge about language are we encoding when we decide to spell “talk” like “tawk”, or “playing” like “pleying”? Some of these variant spelling probably don’t have anything to do with pronunciation, like “gawd” or “dawg”, which I think are more about establishing a playful, informal tone. But I think that some variant spellings absolutely are encoding specific pronunciation. Take a look at this tweet, for example (bolding mine):

There are three different spelling here, two which look like th-stopping (where the “th” sound as in “that” is produced as a “d” sound instead) and one that looks like r-lessness (where someone doesn’t produce the r sound in some words). But unfortunately I don’t have a recording of the person who wrote this tweet; there’s no way I can know if they produce these words in the same way in their speech as they do when typing.

Fortunately, I was able to find someone who 1) uses variant spellings in their Twitter and 2) I could get a recording of:

This let me directly compare how this particular speaker tweets to how they speak. So what did I find? Do they tweet the same way they speak? It turns out that that actually depends.

  • Yes! For some things (like the th-stopping and r-lessness like I mentioned above) this person does tweet and speak in pretty much the same way. They won’t use an “r” in spelling where they wouldn’t say an “r” sound and vice versa.
  • No! But for other things (like saying “ing” words “in” or saying words like “coffin” and “coughing” with a different vowel in the first syllable) while this person does them a lot in thier speech, they aren’t using variant spellings at the same level in thier tweets. So they’ll say “runnin” 80% of the time, for example, but type it as “running” 60% of the time (rather than 20%, which is what we’d expect if the Twitter and speech data were showing the same thing).

So what’s going on? Why are only some things being used in the same way on Twitter and in speech? To answer that we’ll need to dig a little deeper into the way these things in speech.

  • How are th-stopping and r-lessness being used in speech? So when you compare the video above to one of the sports radio announcer that’s being parodied (try this one) you’ll find that they’re actually used more in the video above than they are in the speech that’s being parodied. This is pretty common in situations where someone’s really laying on a particular accent (even one they speak natively), which sociolinguists call a performance register.
  • What about the other things? The things that aren’t being used as often Twitter as they are on speech, on the other hand, actually show up at the same levels in speech, both for the parody and the original. This speaker isn’t overshooting thier use of these features; instead they’re just using them in the way that another native speaker of a dialect would.

So there’s a pretty robust pattern showing up here. This person is only tweeting the way they speak for a very small set of things: those things that are really strongly associated with this dialect and that they’re really playing up in thier speech. In other words, they tend to use the things that they’re paying a lot of attention to in the same way both in speech and on Twitter. That makes sense. If you’re very careful to do something when you’re talking–not splitting an infinitive or ending a sentence with a preposition, maybe–you’re probably not going to do it when you’re talking. But if there’s something that you do all the time when you’re talking and aren’t really aware of then it probably show up in your writing. For example, there are lots of little phrases I’ll use in my speech (like “no worries”, for example) that I don’t think I’ve ever written down, even in really informal contexts. (Except for here, obviously.)

So the answer to whether tweets and speech act the same way is… is depends. Which is actually really useful! Since it looks like it’s only the things that people are paying a lot of attention to that get overshot in speech and Twitter, this can help us figure out what things people think are really important by looking at how they use them on Twitter. And that can help us understand what it is that makes a dialect sound different, which is useful for things like dialect coaching, language teaching and even helping computers understand multiple dialects well.

(BTW, If you’re interested in more details on this project, you can see my poster, which I’ll be presenting at NWAV44 this weekend, here.)

Tweeting with an accent

I’m writing this blog post from a cute little tea shop in Victoria, BC. I’m up here to present at the Northwest Linguistics Conference, which is a yearly conference for both Canadian and American linguists (yes, I know Canadians are Americans too, but United Statsian sounds weird), and I thought that my research project may be interesting to non-linguists as well. Basically, I investigated whether it’s possible for Twitter users to “type with an accent”. Can linguists use variant spellings in Twitter data to look at the same sort of sound patterns we see in different speech communities?

Picture of a bird saying

Picture of a bird saying “Let’s Tawk”. Taken from the website of the Center for the Psychology of Women in Seattle. Click for link.

So if you’ve been following the Great Ideas in Linguistics series, you’ll remember that I wrote about sociolinguistic variables a while ago. If you didn’t, sociolinguistic variables are sounds, words or grammatical structures that are used by specific social groups. So, for example, in Southern American English (representing!) the sound in “I” is produced with only one sound, so it’s more like “ah”.

Now, in speech these sociolinguistic variables are very well studied. In fact, the Dictionary of American Regional English was just finished in 2013 after over fifty years of work. But in computer mediated communication–which is the fancy term for internet language–they haven’t been really well studied. In fact, some scholars suggested that it might not be possible to study speech sounds using written data. And on the surface of it, that does make sense. Why would you expect to be able to get information about speech sounds from a written medium? I mean, look at my attempt to explain an accent feature in the last paragraph. It would be far easier to get my point across using a sound file. That said, I’d noticed in my own internet usage that people were using variant spellings, like “tawk” for “talk”, and I had a hunch that they were using variant spellings in the same way they use different dialect sounds in speech.

While hunches have their place in science, they do need to be verified empirically before they can be taken seriously. And so before I submitted my abstract, let alone gave my talk, I needed to see if I was right. Were Twitter users using variant spellings in the same way that speakers use different sound patterns? And if they are, does that mean that we can investigate sound  patterns using Twitter data?

Since I’m going to present my findings at a conference and am writing this blog post, you can probably deduce that I was right, and that this is indeed the case. How did I show this? Well, first I picked a really well-studied sociolinguistic variable called the low back merger. If you don’t have the merger (most African American speakers and speakers in the South don’t) then you’ll hear a strong difference between the words “cot” and “caught” or “god” and “gaud”. Or, to use the example above, you might have a difference between the words “talk” and “tock”. “Talk” is little more backed and rounded, so it sounds a little more like “tawk”, which is why it’s sometimes spelled that way. I used the Twitter public API and found a bunch of tweets that used the “aw” spelling of common words and then looked to see if there were other variant spellings in those tweets. And there were. Furthermore, the other variant spellings used in tweets also showed features of Southern American English or African American English. Just to make sure, I then looked to see if people were doing the same thing with variant spellings of sociolinguistic variables associated with Scottish English, and they were. (If you’re interested in the nitty-gritty details, my slides are here.)

Ok, so people will sometimes spell things differently on Twitter based on their spoken language dialect. What’s the big deal? Well, for linguists this is pretty exciting. There’s a lot of language data available on Twitter and my research suggests that we can use it to look at variation in sound patterns. If you’re a researcher looking at sound patterns, that’s pretty sweet: you can stay home in your jammies and use Twitter data to verify findings from your field work. But what if you’re not a language researcher? Well, if we can identify someone’s dialect features from their Tweets then we can also use those features to make a pretty good guess about their demographic information, which isn’t always available (another problem for sociolinguists working with internet data). And if, say, you’re trying to sell someone hunting rifles, then it’s pretty helpful to know that they live in a place where they aren’t illegal. It’s early days yet, and I’m nowhere near that stage, but it’s pretty exciting to think that it could happen at some point down the line.

So the big take away is that, yes, people can tweet with an accent, and yes, linguists can use Twitter data to investigate speech sounds. Not all of them–a lot of people aren’t aware of many of their dialect features and thus won’t spell them any differently–but it’s certainly an interesting area for further research.