Here's a useful datum to settle arguments with your spouse

Yes, but who scores better in verbal communication and vocabulary?

Ah, I see -- it just moves the argument to a different plane.

Speaking as a woman, I have no comment...

PZ Myers: "Ah, I see -- it just moves the argument to a different plane."

I wasn't trying to make an argument of it. In my workplace it seems that everything is about equal. However, there are about six women for every man there, so it's hardly the best place to find a representative sample for analysis. When I was growing up ('60s through the '80s), I recall studies (yes, this is anecdotal) that asserted how women were a) better at socializing, b) had greater vocabularies on average, and c) were better at effective communication. I also heard, on more than one occasion, that men were the more active gossips.

Not a move to another plane, just a peripheral query from someone interested in the subject. I'd also like to know if this parallels communication in our closest ape relatives.

There must not have been many Norwegians, Swedes, or Finns. That number is far too high for them. Divide it by 5 at least.

Well, speaking as a man, I have no comment...

This says nothing about the content! ;-)

But seriously, what about the social situations of the participants? Were they at work, say, in offices, or on a building site, driving a truck, or at home with the kids?

The least talkative man uttered 82% of the number of words of the least talkative woman. Now that stretches my credulity.

I have no axe to grind either way, but I am curious about how much the selected population (396 college students) and the sampling method may have affected the outcome. I do not have access to Science to read the whole paper unfortunately.

I would accept the study as more conclusive if it had drawn the sample from across the general population, normalised for age, sex, native tongue, intelligence, etc, etc.

Similarly fixed length samples may give misleading word counts if one sex speaks more quickly than the other. Plus the sampling method presumably could not record the context or setting that the conversations took place in. I would love to see how the word counts stack up in male/male, female/female, and female/male discussions.

b) had greater vocabularies on average,

In the absence of access to the paper, IIRC it has been claimed that it surveys the vocabulary as well, again with no significant differences.

There must not have been many Norwegians, Swedes, or Finns. That number is far too high for them.

Perhaps. Or perhaps it is another factoid as the apparently urban myth that there is a gender difference.

I think differences may come out in cultural contexts - perhaps scandinavians are chatty in different situations.

One factor is seasonal - extreme north/south places makes more people affected by mood swings during winter and spring, usually with a depressing effect. But it isn't scandinavian as such.

Another factor is language speed. Again, while scandinavians may speak slowly, there are others in the same language groups that have similar speeds.

The contexts are random. The subjects -- all of them university students -- were given those recorders, but were not told when the recorders would be switched on and off, and couldn't tell that themselves.

More here and in several other posts in this blog.

Having recently moved to NY and begun riding the subway, it didn't take me long to realize that the loudest and most prolific gabbers were usually teens. This study was done on college-age folk, presumed to be ages 18-21.

I'm not saying that women do talk more than men, but I am saying that if they do, then I suspect this difference would be less pronounced among the young and vital. Do a study on middle aged people. My guess is the results may be different.

Hmm that's one hell of a margin of error. Almost 50% of the stated amount. Not terribly indicative of anything if you ask me.

Herbie- I think that's the spread, not the error on the mean. It just means there's a ton of variability between individuals.

I am probably not understanding something, but the methodology used in the study seems to violate the Nyquist criteria for discrete sampling (2X the highest frequency). On the face of it, if the data is being undersampled, the expected result would be the one given approximately the same word count (or frequency) per gender group (or signal).

Torbjörn, My comment was a joke. It is based on family experience growing up in the upper Midwest. To get an idea, look up "The Prairie Home Companion" by Garrison Keillor.

http://prairiehome.publicradio.org/

Look up the archives for "News from Lake Wobegon".

Speaking as a guy I love how all these guys really want to find a way to put the research into doubt, not even realizing that they are acting on the very cultural prejudices that this study refutes. Sad, but not surprising.

Well, speaking as a man, I find this study to be full of crap. Women clearly use way more words than men, as the following 50,000 word meandering diatribe will prove...

On the face of it, if the data is being undersampled, the expected result would be the one given approximately the same word count (or frequency) per gender group (or signal).

Huh? If I put a colon between "given" and "approximately", your sentence seems to say that the study found "approximately the same word count (or frequency) per gender group". But no, it doesn't. Scroll up and look: it finds a very high variability for each sex, and those ranges overlap so broadly that there is no statistically significant difference between them. The enormously broad ranges don't look like undersampling to me.

I doubt that I speak more than five thousand words on a voluble day.

Must be a lot of blowhards out there spewing hot air.

I second David's recommendation (#10) to look at the Language Log posts on this. Another good starting point there is http://itre.cis.upenn.edu/~myl/languagelog/archives/004691.html, which contains links to lots of earlier posts on gender-difference myths.

If you have any interest in the scientific study of language, or are even just interested in English usage, Language Log is a great place to go. Their fiskings of people supporting stupid restrictions on normal language are fun, as well as salutary for people with a tendency to pedantry (like me).

Language Log is so good that I generally read it even before Pharyngula :)

Jim Roberts

Ah, but who's picking the topics?

At least one study has shown that if men start a topic, women will converse about it; but women must bring up a series of topics until they find one that men will respond to.

So 68% of people utter between about 8,000 and about 24,000 words a day? I'm not surprised at the huge range.
When I live alone, I can go several days without speaking to anyone, except when I buy something at the supermarket and acknowledge the clerk with a "thanks" or say "one, please" to the woman at the Chinese buffet.

16,000 words a day is about 1,000 per waking hour or sixteen per minute. Hard to contextualize that number in any case.

How many words do people in each gender need to convey a concept/ make a point?

It'd be fun to compare idea output/ word input measures. Maybe "talkative" means low output/ input ratio. And to find more explanatory variables than gender-that's a large deviation.

Language Log is so good that I generally read it even before Pharyngula :)

Stupidly, though, it lacks any possibility for commenting, unless you care to hunt up the contributors' e-mail addresses. For comments, you need to go to Language Hat and others listed in the blogroll of Jabal al-Lughat, including the latter itself.

When I live alone, I can go several days without speaking to anyone, except when I buy something at the supermarket and acknowledge the clerk with a "thanks" or say "one, please" to the woman at the Chinese buffet.

I am much the same: retired and live alone(except for the cat, and we don't have many conversations).

There must not have been many Norwegians, Swedes, or Finns. That number is far too high for them.

Not if they had cell phones.

Now why did they do the study with husbands and wives??

Doh, that should have read "Why DIDN'T they do the study with husbands and wives?"

I blogged this study, too.

About 30% of the sample were actually couples, where each was recorded (the Science article did not describe any insights into their dialogues). The rest of the sample was recruited from various Psychology classes and were aware that their speech would be sampled, so there is a potential for significant bias here. Also, they assumed that all participants slept 7 hours per day when making their calculations. (See the above link for the gender histogram of daily spoken words.)

They actually did do a content analysis. There was no male-female difference in the average number of "unique words" (i.e. the number of different words divided by the number of total words multiplied by 100).

Finally, one of the authors made this observation about the semantic content:
"Men talk more about technology, work, money. They also use more numbers," he said. "Women talk more about fashion and about relationships."

I think the number of children a woman has should be taken into account somewhere in here...

Finally, one of the authors made this observation about the semantic content:
"Men talk more about technology, work, money. They also use more numbers," he said. "Women talk more about fashion and about relationships."

How is "technology" any different from "fashion"?

Roy: thanks for the lead to semantic analysis. Will take a look. Btw, today's NYT Mag contains a fun article titled The Gregarious Mind that shows how some kids with Williams syndrome learn how to bond with men simply by being able to talk about sports...

Graculus: I would argue that learning how to use technology is more cognitively demanding that learning how to use fashion...

Torbjörn, My comment was a joke.

Of course it was a joke, but it is also partly an urban myth and partly a genuine observation (I think), so I choose to answer to that. Or maybe I am just too blond. ;-)

Thank you, PZ! I finally ended an argument with my wife by applying peer-reviewed research and logic! How easy!

Now I'm going happily to bed (if only the family dog would just scooch a bit to the left).

No, seriously, my dog reads Coulter.

How is "technology" any different from "fashion"?

Well, if they were talking about actual operational principles and promising avenues of future development, there's an obvious aesthetic-vs.-practical difference. On the other hand, if they were babbling about the year's must-have gadget...

No, seriously, my dog reads Coulter.

Good dog-owners do not allow their dogs to drink anti-freeze, eat feces, or read Ann Coulter.

There's a splendid Arthur C Clarke story about an engineer who runs this experiment in his house, in order to prove that his wife never lets him get a word in (she denies it) - to his surprise, the meter shows that he talks a lot more than she does. He can't believe it - until he discovers a looped tape of his own voice in the bin, which she has been playing to the microphone while he's at work...

My spouse says male psychology/arts students are different from males in general. Result: arguement not settled.

I told my husband all about that study last week; the methodology used, the conclusions, the myths exploded. He said "Yes, dear."

Ahh, the old joke.
How do you tell if it is a Swede that is talking to you.
He's looking at you're shoes instead of his own.
Jag ar glad at du hade ett trevlig resa.
I'm glad you had a good trip.
Thor Bear - sorry about the bad Swedish lettering. I have a US keyboard, and don't know how to convert it. I have a cousin named Torbjorn Larsson. Visited Santa Monica, CA. Went to a Grateful Dead concert in Ventura, CA? Could it be you?

Ken

PZ, thanks for refuting a hypothesis of mine. I had read the media reports on this, and none of them gave the ranges, so I wondered if perhaps men had greater variability, which would explain the stereotype. I guess not, since that variability is itself relatively small.

Ken:

Visited Santa Monica, CA. Went to a Grateful Dead concert in Ventura, CA? Could it be you?

Sorry, not me, I have been in US three times (well, four truly since Hawaii is outside international waters) but never made it to that part of California. Some interesting people and places, nice blend of kitchens, great shopping, bad traffic and no time for concerts - next state, please. :-)

b) had greater vocabularies on average,

In the absence of access to the paper, IIRC it has been claimed that it surveys the vocabulary as well, again with no significant differences.

There must not have been many Norwegians, Swedes, or Finns. That number is far too high for them.

Perhaps. Or perhaps it is another factoid as the apparently urban myth that there is a gender difference.

I think differences may come out in cultural contexts - perhaps scandinavians are chatty in different situations.

One factor is seasonal - extreme north/south places makes more people affected by mood swings during winter and spring, usually with a depressing effect. But it isn't scandinavian as such.

Another factor is language speed. Again, while scandinavians may speak slowly, there are others in the same language groups that have similar speeds.

The contexts are random. The subjects -- all of them university students -- were given those recorders, but were not told when the recorders would be switched on and off, and couldn't tell that themselves.

More here and in several other posts in this blog.

On the face of it, if the data is being undersampled, the expected result would be the one given approximately the same word count (or frequency) per gender group (or signal).

Huh? If I put a colon between "given" and "approximately", your sentence seems to say that the study found "approximately the same word count (or frequency) per gender group". But no, it doesn't. Scroll up and look: it finds a very high variability for each sex, and those ranges overlap so broadly that there is no statistically significant difference between them. The enormously broad ranges don't look like undersampling to me.

Language Log is so good that I generally read it even before Pharyngula :)

Stupidly, though, it lacks any possibility for commenting, unless you care to hunt up the contributors' e-mail addresses. For comments, you need to go to Language Hat and others listed in the blogroll of Jabal al-Lughat, including the latter itself.

Torbjörn, My comment was a joke.

Of course it was a joke, but it is also partly an urban myth and partly a genuine observation (I think), so I choose to answer to that. Or maybe I am just too blond. ;-)

Ken:

Visited Santa Monica, CA. Went to a Grateful Dead concert in Ventura, CA? Could it be you?

Sorry, not me, I have been in US three times (well, four truly since Hawaii is outside international waters) but never made it to that part of California. Some interesting people and places, nice blend of kitchens, great shopping, bad traffic and no time for concerts - next state, please. :-)