Who are language learners? (2021 study of dabblers, polyglots, hyperglots)

Are polyglots really more likely to be gay men?

That’s what this NY Magazine article suggests, as well as a follow-up survey by Alex Arguelles.

But both surveys were extremely small and limited to an elite group of language learners.  They hand-selected candidates out and assumed there was something “special” or “unique” about those groups. Both were more anecdotal evidence than true studies.

So who do these surveys serve?

In March 2021, Relearn A Language launched a community survey for and by language learners

Our goals were:

  1. To determine who within the population of online language learners considers themselves a “polyglot” or “hyper-polyglot”.

  2. To fill in the gaps left behind by the laboratory surveys by reframing questions of gender and sexuality to reflect what is currently known about those identities. The goal was to give a more complete picture of the gender and sexuality identities within the entire community.

  3. To ask additional questions about the motivations and experiences of online language learners–a distinct break from the neuroscientific outlook of previous researchers. Their goal was to find out what makes online language learners biologically or neurologically special, but we want to focus on unique social experiences learners may have had.

  4. To organize groups within the language learning community not by the number of languages mastered (as Fedorenko and Arguelles did), but by some other identity or motivation. The goal was not only to potentially discover a unique pattern within language learners, but to possibly discover which demographics are most underserved by current market offerings. 

  5. Above all, the study would center the agency of individual learners in the community by creating an easily accessible survey for learners of every level; focusing on their motivation and experiences; and independently publishing the data and key findings for free.

After receiving 2,600 responses (well over our goal of 2,500) we closed the survey in May and began analyzing the data.

We also promised we’d release it all once we finished the process.

And as promised.... we have started to release our initial findings!

But before you start reading, you need to know that everything published here–right now–is just the tip of the iceberg.

By late summer 2021 we will have several times as many analyzed questions, as well as a complete academic report about our participants, methods, and results.

Initial Findings (June 22, 2021)

First of all: how did we find all of our participants?

We received over a hundred shares, including by podcasts and conferences, and it broke down like this:

The sources were important because it made sure that, unlike previous surveys, we did not pick participants from a limited pool of online learners.

We also asked which global region participants were from.

This was not only to give us an idea who is in the online community of language learners, but to give us a cultural context.

As you’re going, keep this in mind.

Now let's dive into the first question we wanted to ask: are polyglots that large of a percentage of the online community?

0.8% of respondents labeled themselves as “hyperglots” and 9.1% as “polyglots”.

(We’ll talk about how this interacts with gender soon.)

We can also see that learners who reported to speak 5+ languages (including their native language or languages) only 3.7% of total learners–meaning almost no language learners fall into the categories of “polyglot” or “hyper-polyglot” assigned by previous rearchers.

But how do language learners use these labels?

How many languages makes someone a polyglot? How many languages makes a hyperglot?

We can see that there isn’t a perfect definition from just glancing at the above two charts. But let’s dive deeper.

Here, we find that previous researchers’ use of the word “polyglot” does not fully align with how language learners themselves use the word “polyglot”.

  • 4 people who labeled themselves polyglots also reported that they had “I have some ability in other languages (but not enough to have a full conversation with a native speaker stranger)”.
  • Perhaps more surprisingly, another 84 people who spoke 1-2 “additional languages” (with or without abilities in others) labeling themselves as polyglots (around 35% of all people who labeled themselves as polyglots.)
  • Only 17.3% of people who labeled themselves as a polyglot reported speaking 5+ languages, which is closer in line with the previous researcher’s definition of the word.
Now: are language leaners really more likely to be gay men?

Unlike previous surveys, we didn’t ask if someone was or “male or female”.


  1. Legally, citizens of Argentina, Australia, Denmark, the Netherlands, Germany, Malta, New Zealand, Pakistan, India, and Nepal have the right to a third gender option on federal documents, including passports.
  2. Central to linguistics and language learning, various indigenous people throughout North America recognize additional genders such as two-spirited and muxe–who, in some cases, may be important speakers of minoritized or endangered languages worthy of protection.
  3. Medically, it’s unknown how many humans have genitalia or secondary sex characteristics that are neither male nor female–but it’s estimated that at least 1 in 100.

(And those people are in the online community!)

I’m not sure about you, but when I saw this my jaw fell on the floor.

So why did the other surveys suggest that language learning was dominated by men?

(We’ll investigate that later.

But we also needed to ask about sexuality.

The thing is, we didn’t ask if someone was “homosexual” or not.

There’s so much more to sexual identity than that!

So maybe we’re not exactly likely to be “gay men”…. but language learners certainly don’t seem to be straight men.

In future releases of key findings, we’ll try to find some reasons for why this might be (and why these results look so different than those in previous studies.)

Those were only a few of our findings concerning how we as learners experience life in the online world of language learning.

Now, let’s dive into how our first subsection of learners form a part of the community.

Key Findings on LGBTQAI+ Experiences

After looking at the initial findings posted above, we were left with two big questions:

  1. If men are in the minority, why did they come out as the majority of “polyglots” in the other surveys?
  2. If there are so many people who fall into categories of neither “man” nor “woman”, neither “straight” nor “homosexual”, then why hasn’t anyone else brought attention to that before? 

So we decided to test three hypothesis:

  1. That men are more likely to label themselves as polyglots or hyperglots.
  2. That LGBTQAI+ people feel less represented by the language learning resources currently on the market.
  3. That LGBTQAI+ people feel less a part of the online language learning community.

Let’s see the results.

A Quick Note on Labels

From now on, you’re going to see three labels that we’ll use to refer to different groups of learners.

  • Trans+ is the marker we will use to denote all survey participants who are not cisgender. 
  • LGB+ is the marker we will use to denote all survey participants who are not heterosexual.
  • LGBTQAI+ is the marker we will use to denote all survey participants who are not cisgender and heterosexual.

The reasons we believed these markers were important in the context of this survey is because (1) with the changing and updating of words, there has never been any single agreed-upon definition for umbrella terms within the community, and (2) markers like “not cis” or “not straight / hetro” would center straightness and gender binary when, in fact, we are discussing very different social experiences.

(Readers are also urged to remember that a transgender person may rightfully identify as straight, just as a lesbian or gay person may identify as cis. Both are still welcome parts of the LGBTQAI+ community, just as many participants who fall into any of these three categories may lead heteronormative lives.)

Men+ and Women+ are how we will refer to groups of cisgender, intersex, trans, and/or nonbinary men and women. We will refer to learners who were assigned one gender at birth and maintain that gender as an adult as “cisgender men” and “cisgender women”. We will avoid the stand-alone “men” and “women” when referring to our own data to avoid confusion.

How do different genders label themselves as learners?

In the following graph, we will take a closer look at the labels language learners are using for themselves–this time, by gender.

Since cismen have been the supposed majority, we’ll pay special attention to them.

Note that cis men’s distribution in all categories was relatively even with their overall presence in the community, with the exception of the “hyperglot” category.

Out of 22 respondents who identified as hyperglots, 13% identified as men–consistent with previous researchers’ findings that polyglots are slightly more likely to be men.

But is that self-selection of a preferred identity label, or a reflection of the learners’ actual skills?

We again see men who claim they speak “5+ other languages well enough to have a fluent conversation or read a novel” make up a larger percentage of that category than any other. (51 out of 95 participants.)

This is again consistent with previous researchers’ findings, although all categories are self-reporting and are not indicative of any language user’s actual measured skill in a standardized language.

Examining these two charts, we can conclude that cisgender men do make up a larger proportion of the subgroup known as “hyper-polyglots”.

But why is that?

We can try to find out in the following two questions.

How often do LGBTQAI+ people see themselves represented in language learning resources?

First, let’s establish how often the language learning community as a whole sees itself represented in various language learning materials.

We asked participants to rank how often they felt represented in (1) apps (2) textbooks and (3) audio courses.

Participants were asked to rate each resource

from 1 (“[Resource] always represents people like me”)

to 5 (“[Resource] never represents people like me”).

Participants were allowed to skip a question if they had never used a resource like that.

As you can see, language learners fell neutrally across the spectrum.

But what happened when we asked about podcast and YouTube representation?

Ok… now things are starting to get interesting.

However, because this post is only a small taste of what’s to come, we’re going to examine only the representation in apps next.

We’ll post our analysis of other mediums within the next two months as part of the larger release.

Trans+ learners are somewhat less likely than cis learners to feel regularly represented in language learning apps.

It makes sense on the surface, right?

So the same thing should hold true for sexual orientation–that LGB+ learners would feel less represented than straight learners?

But LGB+ learners reported feeling never or rarely represented 33.9% of the time.

And straight learners reported that they feel never or rarely represented 34.7% of the time.

It’s worth noting, however, 30.0% of LGB+ learners stated they are often or always represented (vs 34.6% of straight learners).

And there are other dynamics at play–since this is an international survey, language learners in regions where LGBTQAI+ attraction or behavior is criminalized may be less likely to be out of the closet, but also less likely to see themselves regularly represented in media.

But the short of it is: we don’t know why this might be.

So let’s see if we can find anything that might shed light on the question in our final segment of this release.

Do LGBTQAI+ learners see themselves as part of the online language learning community?

First, of course, we need to see how all language learners as a group responded.

Considering participants found the survey through community channels, it might be somewhat surprising to see that online language learners tip slightly towards not feeling like they’re part of the community.

Could that be because of how they’re represented? We’ll take a look at that in reports of key findings.

Both cis and trans+ participants reported extremely similar feelings of belonging within the community: largely neutral but lightly tipping towards disagreeing.

  • 24.3% of cis participants disagreed
  • 24.7% of trans participants disagreed

  • 11.8% of cis participants strongly disagreed
  • 12.9% of trans participants strongly disagreed

Trans+ language learners disagree only slightly more than cis counterparts that they are part of a language learning community.

But what about identities within the trans+ umbrella?

Here we can see a more stark difference.

“Trans or nonbinary men” were more likely to see themselves as part of the community than almost any other trans+ identity. (32.2% agreed that they are part of a language learning community.) The few intersex language learners who responded to the survey also felt more positively, although the sample size was small.

Participants who did not select a specific trans+ identity (but did not identify as cis) were by far the most likely to disagree (35.0% disagreed).

Now let’s see if we can find any patterns when looking at sexual orientations.

Curiously, there is no strong difference between how straight and LGB+ sexualities correlate with a sense of belonging.

  • 36.3% of straight participants disagreed or strongly disagreed, although they made up a larger proportion of “strongly disagree”.
  • 36.2% of LGB+ participants disagreed or strongly disagreed, although they made up a larger proportion of “disagree”.


  • 34.8% of straight participants agreed or strongly agreed, although they made up a larger percentage of “strongly agreed”.
  • 33.7% of LGB+ participants agreed or strongly agreed, although they made up a larger percentage of “ agreed”.

Seen together, a learner’s individual identity seems to have more impact on how they see themselves in the community in contrast with whether or not they identify as straight.

  • Asexual/aromantic people were the most likely to disagree or strongly disagree (39.4%) that they were part of the community.
  • Those who are questioning their sexuality are the least likely to disagree or strongly disagree (only 30.2%–making them the only group to disagree less than straight counterparts).
  • Bisexual or pansexual were the most likely to agree or strongly agree that they were part of the community, at 36.3%. (This is at a higher rate of inclusion than this straight counterparts.)
  • “Other… (but not straight) were the least likely of  every group to agree or strongly agree that they were part of the community, at only 27%.

So do we have a lot of answers?

Maybe, maybe not.

But the good news is we have a ton of additional questions we can analyze!

Additional findings on LGBTQAI+ learners, women, and more will be published soon!

We’re excited to give you more data, as well as our full scientific report.

We’ll also be doing a virtual press tour then, so if you represent a content creator then feel free to drop us a comment below and we’ll respond privately via email!

Multilingual Marissa

I'm Marissa, and not only did I fail out of high school Spanish, but I also managed to forget my native language (Polish) in the process. After 10 years of establishing my career in the arts I decided to give language learning one last chance, and many years (and languages) later, I now help others online learn or relearn languages. I now speak English, Spanish, French, Catalan, Portuguese and am relearning Polish, my heritage language. I've also studied Italian, German, and other languages to low- or intermediate-levels for fun. If you want to relearn a language, I'd love to help!

You may also like...

Leave a Reply

Your email address will not be published. Required fields are marked *