Can you figure out how many words are in your vocabulary using samples

[Beamin]

I was thinking how many words do I know. I know it's more then 3 less then 3 billion  but how to narrow it down from there?

Could you take a sample of your writing and see how diverse the words are in it, then compare that to someone/thing else with a known vocabulary and get an estimate? Some how use statistics to do this. Take for example a tweet. That's a small sample size where you can infer things like the general intelligence and education of the writer. Look at trumps tweets: they contain the vocabulary of around a 3rd or 4th grader with an over emphasis of simple adjectives like: "very", "most", "biggest",  or my favorite nonword; "bigly". Since he makes up words and can't Figure Out proper Capitalization you can be quite sure he's around that 3rd grade level. So if we went off that, how many words does a 3rd grader know? Do you think this would be a valid way of sampling? Or you could take a sample of writing and look at the diversity of the wording. But this again would be more qualitative and not quantitative but still would point you in the right direction. I also wonder how many words we will speak in our life times..

[IanB]

Again, nothing whatsoever to do with electronics... 

On the other hand, why don't you take one of the online "test your vocabulary" quizzes? You will get a pretty good answer to your question.

[wilfred]

Take for example a tweet. That's a small sample size where you can infer things like the general intelligence and education of the writer.

Can you? How? I'm more likely to be impressed by good grammar and spelling than sesquipedalian prose.

[GopherT]

I was thinking how many words do I know. I know it's more then 3 less then 3 billion  but how to narrow it down from there?

Could you take a sample of your writing and see how diverse the words are in it, then compare that to someone/thing else with a known vocabulary and get an estimate? Some how use statistics to do this. Take for example a tweet. That's a small sample size where you can infer things like the general intelligence and education of the writer. Look at trumps tweets: they contain the vocabulary of around a 3rd or 4th grader with an over emphasis of simple adjectives like: "very", "most", "biggest",  or my favorite nonword; "bigly". Since he makes up words and can't Figure Out proper Capitalization you can be quite sure he's around that 3rd grade level. So if we went off that, how many words does a 3rd grader know? Do you think this would be a valid way of sampling? Or you could take a sample of writing and look at the diversity of the wording. But this again would be more qualitative and not quantitative but still would point you in the right direction. I also wonder how many words we will speak in our life times..

I think it is a good question but I also think there are several level of research you need to do as you look into it. In addition to your questions, you need to look at some intermediate questions like, "do people use a limited vocabulary when they speak in short, concise sentences, or give short answers."  If so, it will make your quest more challenging. To be honest, I think higher level words tend to have more nuanced definitions so more context is required to use some words (longer or several sentences).

Finally, Tweets are not known for higher-level thoughts or vocabulary.

[IanB]

Vocabulary is generally understood as "How many words do you recognize and know the meanings of?"

How many words you actually use is a matter of how you rate as a writer, and how well you know the readership you are writing for. Intelligent people will choose their words according to the readers. So the words you use are a measure of intelligence.  The words you recognize are a measure of vocabulary.

[IanB]

For instance, would anyone reasonably speak in the following manner? Even if their vocabulary permits them to?

PM: It was the one question today to which I could give a clear, simple, straightforward, honest answer.

CS: Yes. Unfortunately, although the answer was indeed clear, simple, and straighforward, there is some difficulty in justifiably assigning to it the fourth of the epithets you applied to the statement. Inasmuch as the precise correlation between the information you communicated and the facts insofar as they can be determined and demonstrated, is such as to cause epistomological problems of sufficient magnitude as to lay upon the logical and semantic resources of the English language a heavier burden than they can reasonably be expected to bear.

[legacy]

Intelligent people will choose their words according to the readers.

this might imply that someone judges a non-English speaker similar to a ten years old retarded kid.

This expression "ten years old retarded kid" was formulated by an American dude living in Virginia with the precise intent of making the point about what he looks at in a sentence to recognize language skills of intelligent people, and, since my vocabulary doesn't allow me to choose words appropriately for the readers, he concluded that.

I do find irritating the definition of intelligence, in the first place: he sucks at math, and his competences in computer science are really embarrassing, for example, but he is very good at writing urban dictionary with a lot of "colored words" (so he defines what he says)

[Buriedcode]

I would be weary of anything that appears to, or claims to measure "intelligence" as it isn't something that is particularly easy to measure.  IQ tests and the like are quite good at gauging education level and and often used as a form of metric (to compare different demographics), but it is also controversial because it implies ones intelligence is based on not only level of education, but *where* they were educated, rather than some intrinsic ability.

Ultimately, in this day and age, anything that claims to measure your intelligence will be biased to show a greater intelligence, in the same way "vanity sizing" affects clothing.  This is why online IQ tests almost always say 130-150 IQ even when provided with incorrect answers.

With that said.. are you trying to measure "intelligence" ? Or something else?  You seem to make a lot of leaps and assumptions that might not have any basis in fact.

Considering there are several free-of-charge "always-on" voice recognition facilities, google, Alexa, Cortana, Siri etc.. and these can recognize an individuals voice, maybe there is an app that can record everything you say for statistical purposes. It is much easier for writing - just copy/paste every email, forum post, or search, into notepad.

Although its probably dangerous to comment on political figures, I agree about Trump, his tweets look like they were written by a simpleton.  That might not mean that much though.. I'm sure I have sent plenty of tweets/texts that were dumb.

[apis]

I would be weary of anything that appears to, or claims to measure "intelligence" as it isn't something that is particularly easy to measure.

And even harder to define.

IQ tests and the like are quite good at gauging education level and and often used as a form of metric (to compare different demographics), but it is also controversial because it implies ones intelligence is based on not only level of education, but *where* they were educated, rather than some intrinsic ability.

IQ test only measure how good you are at IQ tests. The reason why people sometimes use them is that there is some correlation between IQ tests and income and some other metrics (it's fairly weak, but it's significant as far as I'm aware), but it's not really correct to call it a measure of intelligence.

[IanB]

I do find irritating the definition of intelligence, in the first place: he sucks at math, and his competences in computer science are really embarrassing, for example, but he is very good at writing urban dictionary with a lot of "colored words" (so he defines what he says)

I think you are missing my point. I think intelligent people will try to avoid colorful and flowery language in favor of being simple, direct and to the point.

If some moron in America thinks this is the mark of a 10 year old retarded kid, well it becomes very clear who is not intelligent.

[Nominal Animal]

I posit that the quality and complexity of language used has no causal link to intelligence (or more specifically, the g-factor).  My evidence: Up Goer Five and Thing Explainer by Randall Munroe.

Kidding aside, individuals have huge differences in their spoken and written language skills.  This is because they involve different parts of the brain.  They are also rather specialized/localized, in that you can lose your language skills due to brain damage, without losing (much) of your intelligence.

On Twitter and elsewhere, those able to express themselves using complex language often simplify their output, to avoid being seen as snobbish -- because the prevalent style is simplistic, with superfluous errors thrown in.  (Conversely, many utterly stupid people learn specific jargon, well enough to fool others, and think that makes them intelligent and worth listening to. )  Personally, I would not make the mistake of assuming someone is a twit just because some of their tweets use language that a three-year-old understands; especially if they are clever enough to win in a presidential election.

Note that that is not a comment about Trump alone: just compare the pre-Presidential and Presidential speeches by Obama and Bush.  You'll see the very same simplification in their output.  (I once talked about this with my friends, and wondered whether it is due to CIA spiking their food/drink to keep them complacent...  The difference is that striking, especially for George W. Bush.  He used to be eloquent before his precidency, no matter what you think about his politics.)

Of course, English is not my primary language, and me fail English often (especially as I use predominantly written, not spoken English).  So this is just my understanding of the matter.

[IanB]

I posit that the quality and complexity of language used has no causal link to intelligence (or more specifically, the g-factor).

What if someone is aware of who they are speaking or writing for, and carefully chooses their words to communicate most effectively? Or conversely, if someone seems blissfully unaware and uses all the wrong words? Is that not a marker of intelligence (or lack of)?

(Example: Someone in a foreign country speaks their own language loudly and slowly in the assurance that the listeners must understand them if they speak slowly enough, while making no attempt to communicate in the local language used by the listeners.)

[Nominal Animal]

I would be weary of anything that appears to, or claims to measure "intelligence" as it isn't something that is particularly easy to measure.

It is more of a problem of how "intelligence" is defined, really.

In psychometrics, they use the g-factor.

In larger scale statistical analysis of IQ tests, it looks like there are at least three aspects to "intelligence": reasoning (logic), short-term memory, and language/verbal ability.

It looks like in general, IQ tests are statistically reliable.  That is, testing a single person using a single test should be treated as a single sample, and not a reliable indicator.  However, applying several different tests should yield a more reliable picture of the general intelligence or cognitive abilities of that person.

(The way the tests are constructed, however, is such that each result is likely to be too low than too high.  You'll see that in multiple-choice questions there are at least six options to choose from, to minimize the chance that picking choices by random gives a positive result.  Then, there are more complex evaluation methods, similar to those used in psychological tests, where the combinations of choices are evaluated, instead of each question as a singleton.)

So, it's not that the IQ tests themselves are iffy per se; it's just that one must be critical of how to interpret the results.

And yes, this does mean that anyone who is a member of Mensa, does not understand statistics and/or psychometrics.

What if someone is aware of who they are speaking or writing for, and carefully chooses their words to communicate most effectively? Or conversely, if someone seems blissfully unaware and uses all the wrong words? Is that not a marker of intelligence (or lack of)?

I am socially inept: I tend to always communicate the same way.  Does that reflect my lack of intelligence, or is it just a quirk of my personality?

My view is that social awareness of context, and applying it to communication, is cleverness -- a skill, basically, that many learn as a kid -- rather than an indicator of intelligence.

(Example: Someone in a foreign country speaks their own language loudly and slowly in the assurance that the listeners must understand them if they speak slowly enough, while making no attempt to communicate in the local language used by the listeners.)

I've known one such person, decades ago.  They were traveling outside Finland for the first time in their life.  The tourists they had met in Finland had all knew a phrase or two of Finnish, out of a frase book I think (and speaking slowly and loudly and enunciating clearly of course helps then).  They weren't stupid, just very, very inexperienced: they didn't know any better.

On the other hand, even though they were old, they did learn key phrases quite quickly...  like "Una cerveza, por favor."

So, still no from me: I think that is a social issue, nothing to do with intelligence per se.  (Wrt. americans, inasmuch as that actually happens, I blame the perception of American Exceptionalism.  Otherwise just us hicks let outside for the first time.)

[helius]

There is a site that estimates your vocabulary size: http://testyourvocab.com/
It asks you to answer truthfully if you are able to define each word (so there is nothing that stops you from cheating).

[Domagoj T]

I got 22 100.
Sounds a bit high for non native speaker. 

[IanB]

I got 22 100.
Sounds a bit high for non native speaker. 

It told me (native speaker) 35,700 words.

There were many words I recognized, but could not adequately define. In the spirit of the quiz, I did not tick the box for those.

There were many more words I have never seen before and might have guessed that someone made them up.

I continue to be amazed at how many words there are in the English language.

[Beamin]

For instance, would anyone reasonably speak in the following manner? Even if their vocabulary permits them to?

PM: It was the one question today to which I could give a clear, simple, straightforward, honest answer.

CS: Yes. Unfortunately, although the answer was indeed clear, simple, and straighforward, there is some difficulty in justifiably assigning to it the fourth of the epithets you applied to the statement. Inasmuch as the precise correlation between the information you communicated and the facts insofar as they can be determined and demonstrated, is such as to cause epistomological problems of sufficient magnitude as to lay upon the logical and semantic resources of the English language a heavier burden than they can reasonably be expected to bear.

Sometimes I would talk like that to make people feel stupid, especially when they were trying to sound intelligent.

But good point. When I speak to foreign people I will speak like they do; Asian people I won't use plurals, or when dealing with normal people I will "dumb down" what I'm saying, not only so they will understand me, but so I don't come off as arrogant, or make them feel dumb. You would be surprised how even reasonably intelligent people, like doctors (doctors especially who are supposed to know more then the patient and certainly used to being the "smartest guy in the room"), will cut you off when you start to go over their level of understanding. You can try this out; I can think of at least three clear examples where I got the same outcome from reasonably intelligent people: start talking about neuroscience as it relates to their or other peoples behaviors and use the anatomical names. I always get the same reaction where they cut me off and say something like "well we won't go there", or "I don't know anything about that". But average people will say "wow you are really smart" or give a positive response. It's called "intellectual intimidation", and I had the hardest time with this when I was younger and dealing with professors or doctors, that would say something wrong and true to Asperger's form I would blurt out the correction. Learned that the hard way as I was graded harshly in courses I was taking, or had doctors get up and leave the room.   

I bet I lose points to my low vision and relying of autocorrect and spell check to fix my writing. The spell check has the vocabulary of a 4th grader and often use Google's "Did you mean..." as my spell check if I'm not being lazy.

I  bet though you could use statistics and an index of knowns, like known words, and feed an algorithm to spit out an answer. I'm sure google does this as a metric when it reads our Gmail accounts to sell to advertisers. If a group of people are really dumb, or believe anything they are told, you could sell snake oil to them.

[Buriedcode]

IQ tests and the like are quite good at gauging education level and and often used as a form of metric (to compare different demographics), but it is also controversial because it implies ones intelligence is based on not only level of education, but *where* they were educated, rather than some intrinsic ability.

IQ test only measure how good you are at IQ tests. The reason why people sometimes use them is that there is some correlation between IQ tests and income and some other metrics (it's fairly weak, but it's significant as far as I'm aware), but it's not really correct to call it a measure of intelligence.

Indeed.  Technically as its a quotient, ones IQ is also related to the scores of those in the same "group" you're measuring (since scores should be normalized to 100).   The fact its a relative "test" is often played down, and is taken to be something that is absolute.  For the record, just in case I gave a different impression, I dont' believe it is a particularly good measure of intelligence, and often the only people who care about it, don't really understand what it is used for.

I might try that test, but as I mentioned, I'm weary of online tests - often they don't really tell you anything except that everyone is great.

[IanB]

I might try that test, but as I mentioned, I'm weary of online tests - often they don't really tell you anything except that everyone is great.

If you are referring to the vocabulary test above, do try it. It's fun. What it I learned from it was utter amazement that there could be so many words I have never seen or heard before that apparently are real words in the dictionary 

[ucanel]

...
So the words you use are a measure of intelligence.  The words you recognize are a measure of vocabulary.

I agree.

English is not my mother tongue and
i even have not live any English spoken place.

Before i read your post
i will say ' I understand easily but could not talk English well'
after reading your post;
my vocabulary of English is good but
my intelligence of English is poor.


Also i need to point out that this is a good subject for so many aspects.
I was watching a tv program about artifical intelligence and
then they talked about the language as an example of self learning process.
Thay also said there is an experiment about learning language while sleeping,
before this my opinion about sleep learning was
'if you try to learn language while sleeping,
you would speak only in your dreams'
but at the experiment they gave the subjects
grammerically correct order sentences of an unexisting language (experimenters created a language)
and when the subjects were awake
they asked the ordered and unordered sentences to identify which one is which
and the subjects could seperated the wrong and true ones.

[legacy]

I think you are missing my point. I think intelligent people will try to avoid colorful and flowery language in favor of being simple, direct and to the point.

My point is that you cannot talk about "intelligence", you can only talk about "language skills".

[CatalinaWOW]

I think using a persons writing as an estimator of vocabulary is a fools errand.  In addition to the previously mentioned tendency of wise people to write for their audience I will give you my personal situation.  I read a great deal and encounter and learn a great many words in this reading.  Words I will never use in conversation or in my writings.  So they are part of my vocabulary (words I understand), but not part of my vocabulary (words I use).

Just a random sampling of some of these understood, but unused words.  Fubsy.  Cringle.  Collop.  Hod.  Orlop.  Ort.

Why don't I use these words and their brethren?  Several reasons.  Some are professional terms in areas I am not involved in.  Some could be useful, but I know they would at least cause a pause for explanation or dictionary lookup.  That didn't bother the authors of the books where I encountered these words, but I write to communicate, not to impress.  And finally, they aren't in the most rapid access part (cache memory so to speak) of my brain so other, more common words get used.

[legacy]

Oh, it's called "artificial intelligence", NLP(2) is a specific application: if a toaster is able to insult you in fluent English, then is it more intelligent than a human being?

human being: hey thing? I want two slices of toasted bread,  faster than a fsck
It, the *thing*: dear "old skin"(1) I won't serve any human being unable to show intelligence
human being: oh, shit! Do it, or I will smash you, stupid thing
It, the *thing*: is it the best you can do, skin? some grunts like a pig?
human being: I want my goddamn toasted bread!
it, the *thing*: do it by yourself, skin

a laughing that is not a human laughing
then it, the *thing*, underlines: "if you are able to do so!"

human being: what? the? fsck? is your problem, stupid thing?
it, the *thing*: precisely!
it, the *thing*: I have a ROM with zillion billion words
it, the *thing*: and I can correlate them to the audience more clever than a human being
it, the *thing*: while your vocabulary not only is reduced but only able to some grunts like a pig.
human being: so what?
it, the *think*: this means I am more intelligent than you, and serving an inferior intelligence is illogical


(geek humor  )


(1) it's how a machine has decided to call a "human being"
(2) NLP, natural language processing ... e.g. Amazon Echo, aka "Alexa"