Speed of speech and its implications
The NYTimes decided to report on the extent to which Hungarians are better than Americans at recalling store prices. Given that most blogging I do about Hungary seems to result in a discussion of the Hungarian language and given that the authors explain the findings based on language differences, I thought I’d take this opportunity to address the issue head on.
Let’s start with the findings:
Hungarians are far better than Americans at recalling long prices; on average, they can recall 19 to 24 syllables with decent accuracy, while Americans can recall only 13. The authors suggested that this was because Hungarians speak 41 percent faster, both out loud and when repeating sounds to themselves “subvocally.â€
The NYTimes piece ends right there. That’s not fair, the author left out the most interesting part: how do we know how fast Hungarians speak in comparison to Americans?
First, let’s set aside the question of who these “Hungarians” and “Americans” are based on the study. * (I actually tend to care about these things, but I don’t want to derail the speech speed mystery.)
I looked up the article [subscription required] upon which the NYTimes blurb was based. It appears in the September issue of the Journal of Consumer Research. After presenting the findings, the authors suggest the following:
One possibility is that Hungarians speak faster. Research on the memory span across different languages has suggested that languages that use more syllables on average per word tend to be spoken at a faster rate, possibly to attain “an optimum rate of information flow … to suit the human capacity for message transmission and reception” (Naveh-Benjamin and Ayres 1986, 749). Because we did not find any published work on speech rate or language “length” that included Hungarian participants, we compared the pronunciation speed between Hungarian and American participants for the prices used in the previous studies to test the speech rate explanation. We asked 10 people in each country to read subvocally three samples of the prices used in the experiments. The total pronunciation time was then divided by the number of syllables, which resulted in an average subvocal reading time per syllable of 0.198 sec. for the Hungarians and 0.279 sec. for the Americans. Thus, the Hungarians clearly are faster speakers [..].
So we know that Hungarians speak faster than Americans based on ten people each? Are all Americans’ pronounciations and speech similar enough to generalize? Hungarian doesn’t have that much of a dialect, but I think people’s speech speed still differs. (I know, I know, I couldn’t really set aside the issue of sampling after all.)
So does speed with one language translate into speed with another one? As a native Hungarian speaker, would my speed in English or French be different from native speakers of languages that are, presumably, spoken slower?
By the way, I’m not convinced by the idea that languages with longer words (longer = higher number of syllables that is) are spoken faster necessarily. How does this then explain variation in speed within a language. Swiss romand folks speak French – pretty much the same French – slower than their neighbors in France. How do we explain that? Of course, this is not my area and am not familiar with the cited 1986 study, but I am a bit skeptical. Interesting…
So this is supposed to explain why Hungarians recall longer prices better. Personally, I don’t think it would matter to me how many syllables there were in a price on a tag. I tend to be way too visual for such details. I’d recall the image of the number not its name. Or does it not work that way? Hmm… I’ll have to go test it next time I’m in a store.
Mark Liberman of LanguageLog has also commented on this story here and here. Moreover, he has decided to run his own language speed experiment on native Hungarian speakers. If you are one, please send him a note an participate. The original peer-reviewed journal article based its findings on just ten participants, let’s see if he can get at least as many.
[*] I think the Americans are less than 100 college students. The sampling for the Hungarians wasn’t as obvious to identify and I am not about to go through the original study in detail. If anyone figures it out, do post it in the comments.
August 18th, 2006 at 9:13 am
In reply to your comment:
“As a native Hungarian speaker, would my speed in English or French be different from native speakers of languages that are, presumably, spoken slower?”
Speaking from experience (being your brother and working at a University) I have to say that you speak English faster than most of my co-workers or students. I have also noticed that when I get “excited” about a topic that I am covering in class I tend to speak much faster in English, at about the speed we use in interesting Hungarian conversations.
August 20th, 2006 at 7:31 pm
Your point about the sample size, although you hesitate to dwell on it, is completely valid. I’m a native English speaker, fluent in Hungarian (OK, rusty), and I know fast talkers and slow talkers in both languages. The notion that “Hungarians speak faster” is preposterous to me.
In fact, if anything, when it comes to television and radio, Hungarians speak more slowly and far more clearly than Americans.
The idea that Hungarians are better at remembering numbers is intereseting though.
August 20th, 2006 at 7:35 pm
Michael, you raise an interesting point when you mention the issue about numbers. Obviously this has been about numbers all along, but I haven’t focused on it much. Hungarians are ridiculously overrepresented in math and related fields among top scientists. Some have suggested that this is at least in part related to the Hungarian language. If that is the case then perhaps it is easier for Hungarians for some reason to remember numbers. I wonder if they would do similarly well recalling other types of information.
August 21st, 2006 at 12:19 am
When people ask me how I learned such a “hard” language as Hungarian, I bore with them examples of how logical it is; how, if you’re willing to accept some basic differences in how sentences are put together, it actually makes a lot more sense than many other languages. (It’s wonderful; highly recommended 😉
Specifically, I wonder if this ties into the whole numbers question: To build a sentence in Hungarian, often times, you really need to have formed the whole idea in advance, and so you do need to keep a lot of information in your head before completing the sentence. (To thoroughly explain what I mean would be inappropriately long for a blog comment, I’m afraid.)
Eh, maybe that’s a bit of a stretch. (shrug)
August 21st, 2006 at 12:25 am
Oh yeah, I had forgotten to comment on the fact that you speak Hungarian. That’s very impressive. I always find that curious.. why would someone learn that language? I don’t mean to suggest that it’s not a neat language, I think it’s great, I just question its utility for non-Hungarians.
Being a native speaker of the language, I have never thought about it the way you talk about it in your comment, Michael. I’d be curious in the more thorough explanation.. I don’t have a cut-off for blog comments. (Well, this field might, but then you could use two comment fields one after another.:)
August 21st, 2006 at 1:28 pm
I lived in Budapest between 1989 and 1992, studying at the Liszt Academy of Music (but mainly enjoying a lot of cake). So that’s how I learned the language. I do what I can to maintain it, but it’s gone really rusty now.
Anyway, OK, I’ll give it a shot, but it might be easier to explain in the context of reading a sentence, rather than building one in conversation. Let’s take a look at this sentence (via index.hu):
“Az augusztus 20-i tűzijáték közben kialakult, négy halálos áldozatot követelÅ‘ katasztrófa felelÅ‘seit keresÅ‘ kormányzati vizsgálat megállapÃtotta, hogy a rendezvényt szervezÅ‘ Nexus kft. mindent megfelelÅ‘en csinált, viszont azt érdemes volna kivizsgálni, hogy a meteorológiai intézet figyelmeztetésének miért nem lett foganatja.”
Just looking at the first clause, there’s a ton of information that English would require breaking up into smaller sentences. Here’s an attempt at packing this all into one English sentence.
“A government investigation into who’s to blame for the catastrophe that came about during the August 20 fireworks, causing four deaths, determined that…”
The subject of the sentence is “investigation”, and that needs to be at the beginning of the sentence in order to make sense in English. In the Hungarian, it’s way downstream right before the verb “determined” at the end of the clause.
So, in the Hungarian, you have to get through all the descriptive stuff before knowing what the subject is.
I had to really pick this apart to figure out what it means:
– Az augusztus 20-i tűzijáték közben kialakult,
(brought about during the August 20 fireworks)
– négy halálos áldozatot követelÅ‘ katasztrófa
(catastrophe causing four deaths)
– felelÅ‘seit keresÅ‘
(searching for those responsible)
– kormányzati vizsgálat
(government investigation)
You don’t know where the sentence is going until it’s there already. As a non-native speaker I find this really exhausting. So, I wonder if a lifetime of hearing and reading everyday language does indeed strengthen ones capacity to store these little details.
Does any of this make sense?
August 21st, 2006 at 2:43 pm
Thanks, Michael, this is very helpful, I see now what you meant. To be fair, that sentence structure from a newspaper is not exactly the type one would use in day-to-day conversation, but you’re right that it is something most Hungarians would understand. There are such daily language/newspaper language discrepencies in lots of languages though, like Germany and I’m pretty sure Japanese (just to name two random ones only b/c I have some experience with both). In any case, it’s helpful for me to see it broken down in that way, thanks. It would be interesting to know how this translates to memory.