A fascinating AskMetaFilter thread set off by the question “Why didn’t I say anything until I was three?” It’s full of interesting anecdotes and references to studies, and one commenter quotes this great Einstein story:

Otto Neugebauer told the writer the following legend about Einstein. It seems that when Einstein was a young boy he was a lake talker and naturally his parents were worried. Finally, one dat at supper, he broke into speech with the words “Die Suppe ist zu heiss.” (The soup is too hot.) His parents were greatly relieved, but asked him why he hadn’t spoken up to that time. The answer came back: “Bisher war Alles in Ordnung.” (Until now everything was in order.)

I think I once knew how old I was when I started talking, but I’ve forgotten. Anyway, late talking doesn’t seem to be a cause for concern.


  1. Christ, I Need A Drink says

    I’ve heard that story before, and while I love it it’s always struck me as far-fetched. Could a story like that really be true?

  2. I’m not sure the Einstein story is all that unusual. My brother-in-law supposedly also didn’t say a word until he was 3 and then immediately began speaking in full sentences and has been a bright well-adjusted individual ever since, very succesful in fact. I’ve always been told that late talking is, in fact, something to be concerned about in general – it is often an indication of autism or Asperger’s (and Einstein’s case tends to support that rather than disprove it).

  3. Isn’t that just a reworking of the story of Lord Macaulay?

  4. The Log took this up in 2007: we get both Macaulay and Einstein as well as Carlyle (“What ails wee Jock?”), which last sounds more convincing to me. I accept that kids who speak late, speak age-appropriately from the start, but the Einstein, and surely the Macaulay, sound way off even for a very bookish child.
    But my favorite version is an apocryphal kid whose first words are “Please pass the salt”, with the usual parental followup. His reply: “I always sat next to the salt before.”
    At 16 months, my grandson is still only babbling, with zero semantic content, and a teacher is coming today to evaluate him to see if he needs early intervention. (A speech therapist has already seen him and concluded he doesn’t quite meet the State’s criteria and must wait another three months for therapy.) He has recently mastered coda consonants, though.

  5. My 5 year old grandnephew talks all the time and drives everyone crazy. Count your blessings.

  6. Jan Freeman says

    I dimly remember from grad school that late talking/development was a trait of some heroes of early legend, who were thought to be delayed before their sudden blossoming — Beowulf perhaps among them?

  7. Too late to comment at Language Log, so I’ll hijack the thread and point out that Russell’s description of his (explicitly) fictional Patagonians quoted there looks like a muddled reminiscence of something he read (or had explained to him orally?) about polysynthesis with incorporation.
    And while I’m at it, good ol’ Malinowski, with his Trobriand Islanders whose metaphysics recognizes only masses, not individuals? Well, it turns out that Kilivila is a perfectly straightforward classifier language, just like Chinese or Burmese; so it’s true in one sense that all their nouns are mass nouns, but metaphysics? Nah.
    (On the other hand, “white horse not horse” wouldn’t be such an important philosophical point in any language but Classical Chinese.)

  8. Chad Hansen’s first book on Chinese philosophy did derive a metaphysics from the Chinese mass nouns, though perhaps john C. already knows that.

  9. Boris van ‘t Blad, our son and heir, has a month and a dag to go until his second birthdag and it remains to be seen if the dreaded Kinderkontrolle will accept his verbalities as reaching their exacting standards at this milestone.
    Stupid Kinderkontrolle.

  10. Some of us that have a low IQ, wait until we can say a short Sentence, my father always assumed I was stupid because I refused to use the usual words Papa and Mama, and my first recorded utterance was “Damn it, I broke it”.
    We have bench marks, but Nature [DNA] likes to keep all Scientists on their toes by not keeping to the rules that Homo Sapiens/erectus expect.
    In the animal kingdom it be called sports.
    6. a. A plant (or part of a plant), animal, etc., which exhibits abnormal or striking variation from the parent type, esp. in form or colour; a spontaneous mutation; a new variety produced in this way. Cf. SPORT v. 8, sport of nature at Phrases 4a.

  11. All the late-speaking children cited here are boys. May I remind the learned audience that grown men are well-known for not wanting to speak much, at least about emotional matters? Perhaps some boys only start to speak when confronted with a technical hitch, or a problem that needs to be solved: “Damn it, I broke it”, or “Please pass the salt” to resolve the salt-at-a-distance problem.
    Are late-speaking girls an unusual phenomenon?

  12. marie-lucie says

    It seems that the same story (with different words) is being told about a number of famous men (yes, all boys).
    I think it is common (though by no means universal) for boys to start speaking much later than girls, but then to start speaking much better than expected, as if they had been rehearsing speech in their minds before actually uttering any words. So as long as the little boy understands speech, communicates non-verbally and otherwise behaves at a level appropriate for his age (which rules out deafness and retardation), I don’t think there is anything to worry about until about age three. Many otherwise intelligent boys don’t start speaking until two and a half or so, while most girls speak by 18 months but go through the various stages and at first are not understandable except by their immediate family.
    Last summer I attended a lecture by a man who studies children’s pre-linguistic abilities (what they can do before they start to speak). Among other things he showed a video of a non-speaking boy who looked about two+ years old, faced with various experimental situations, such as if a person left the room where the child was, and came back after some object had been introduced or hidden in her absence. The key point was that the child not only noticed the difference but was obviously eager to make the returning person notice it too (but not the people already in the room, who had witness the event) . The experimenter’s conclusion on the results of similar experiments with a variety of children and situations was that human beings want not only to communicate their own needs (as crying babies do) but also, very early, to share experiences, something that other primates do not do, and this is what he thinks is the basis for the origin and development of language. I suppose that this means that children start speaking, not only around the time when they are becoming physically separate from their parents (by being able to walk) but when gestures are no longer sufficient for them to attract other humans’ attention and transmit their experiences.

  13. I love that story. Here’s another far-fetched but true one: there is a German novelist called Hanns-Josef Ortheil ( who didn’t speak till he was 7, because neither of his parents spoke. His mother stopped speaking out of shock when one of her childred died (she had two still-born and two died in the war, and Ortheil was the fifth). He was regarded as stupid at school. The first sentence he said was ‘Gib mal her’ (‘Give it here’) when he was playing football with other children and wanted the ball.

  14. marie-lucie says

    I read a report about another boy who did not speak because he was not spoken to. His parents were Italian immigrants to Canada, and the husband went to work either in mining or lumber camps, so that he spent long periods of time away from his wife and child. The wife learned almost no English and did not have any friends, but she thought that being in Canada she should not speak Italian, so she did not speak to her son, although she was otherwise a good mother. The father spent too little time with his family, and that at long intervals, to speak much to the child, who did not play with other children. At the age of 5 or so he went to school, where for the first time he was surrounded by children and adults who spoke to him. I don’t know what happened to him later, but it is considered almost impossible for a person older than about 5 years old to acquire a first language – the person will learn to speak some, but never like a native speaker (it is different for a second language learned in childhood). So the story of Ortheil seems to be exceptional. Perhaps no one spoke in the home, but he heard other people speak, enough to gain a passive knowledge of his language.

  15. I just thought I’d second marie-lucie’s comment that in the absence of evidence for deafness or other developmental delay, late talking isn’t too worrisome.
    m-l, I’m not sure about your comment that “it is considered almost impossible for a person older than about 5 years old to acquire a first language – the person will learn to speak some, but never like a native speaker”. As I understood it when we touched on the topic of a ‘critical period’ in language acquisition, the eventual competence drops fairly continuously with delayed exposure to language from about birth through puberty, at which point the eventual competence levels off. The evidence I’m thinking of comes from deaf children of hearing parents, who were not getting any linguistic input they could benefit from until they entered a school for the deaf, where sign language was used and taught.
    I can probably dig that reference up if I try… it was an interesting paper. 🙂

  16. marie-lucie says

    CL, I don’t have a reference at hand for my comment, but I didn’t invent it! The “critical period” seems to be different for first and second language acquisition (but there is far less evidence for the first language than for the second, since it is fortunately exceptional for children to be raised without language at all).
    As to deaf children of hearing parents, apparently families with a deaf child tend to make up a rudimentary form of sign language, specific to the family, so there is some kind of linguistic input, although not as versatile as fully-fledged sign language.

  17. Of course you didn’t invent it! There are a lot of opinions out there about how this works. I just thought I’d put my two cents in about the current state of the literature, since it came up in my class last semester.
    I think the argument for using deaf children of hearing parents as a test case for the critical period hypothesis is that the crucial part of language exposure is its structural richness. Home-sign systems don’t tend to have much consistent structure, so even if these children are getting a little input, the reasoning is that it’s not rich enough in the right ways, and doesn’t count.
    The nice thing about looking at these kids in comparison to other populations in the literature is that they don’t have a clear first language, which could either interfere or help, nor do they have any of the severe deprivations that can go along with late language exposure and that might themselves cause later difficulty (e.g. Genie).

  18. marie-lucie says

    Well, CL, I don’t claim to be up-to-date about the current state of the literature, so perhaps you can suggest something I could read.

  19. My parents were worried that I didn’t speak until I was two. My first word was “Paiper!”, an imitation of an Australian newspaper seller. As I went on to become a journalist, that may have been significant …

  20. This is a good, (relatively) recent chapter on the critical period hypothesis.
    I hope I haven’t offended. I’m afraid I often end up with my foot in my mouth (keyboard?) when I try to chime in in the comments here. :-/ Maybe with practice I’ll improve?

  21. This one is less recent, but focuses more on the evidence from language acquisition.

  22. marie-lucie says

    CL: I hope I haven’t offended. No, not at all! I will read the articles carefully. Thank you for finding them!

  23. no problem!

  24. few children who do not have any words by 3 turn out to have normal language skills.
    not having any words by 18 months is a “red flag” for possible problems—one of which is hearing loss (which frequently can be corrected). such children should at the very least have a hearing test.

  25. few children who do not have any words by 3 turn out to have normal language skills.
    not having any words by 18 months is a “red flag” for possible problems—one of which is hearing loss (which frequently can be corrected). such children should at the very least have a hearing test.

  26. michael farris says

    There seems to be some evidence that the ‘critical period’ applies more to speech than sign. Some years ago I read a book (can’t remember enough to cite the title) on languageless deaf adults in California (prelingually deaf illegal immigrants who never learned to speak and who did not know any particular sign language either). The author was an ASL interpreter who worked with one such languageless adult. He did eventually learn more or less fluent ASL (though his signing remained less ….natural(?) looking than that of other deaf adults.
    I wonder if the critical period has less to do with qualities of the brain and more with the fine motor skills needed for speech .

  27. Per Jørgensen says

    There are times (many) I wish my children and step-children hadn’t learned to speak quite so much.
    There are also times (many) I wish my children and step-children weren’t quite so eager to share every minute experience with me.
    There are times (many) I envy parents of children who would just wait until they really, actually, and legitimately needed something. Like Herr Einstein. Why can’t my children be more like him?

  28. Per, just yesterday I described my adorable and beloved grandnephew as a machine-gun battery of jabbering. He wore out his ma, his pa, his grandma, and me.
    It’s at age 5 that that seems to begin. Younger kids try out what they’re saying with charming tentativeness, but when they get they’re confidence up it’s full speed ahead for many years.
    “Grandma! You’re not listening to me!”

  29. marie-lucie says

    MF: I wonder if the critical period has less to do with qualities of the brain and more with the fine motor skills needed for speech .
    This seems to be true for second language acquisition, since the right phonetics and even phonology are very difficult to acquire past the critical period, even if the rest of the language is near perfect. But we are talking about first language acquisition: persons who for one reason or another did not acquire language as small children have great difficulties not just with pronunciation but with grammar (word order, endings, etc).

  30. Per "Barba Roja" Zarate y Jørgensen, a.k.a. Mr. P says

    Mr. Emerson, I can sympathize. Our four jabber batteries like to lay covering fire by lighting up all at once. Shock and awe. As soon as I step out of the kitchen into the living room, I look and feel like Jon Voight’s character at the end of “Runaway Train,” face first against the headwind on top of the locomotive.
    The human tendency to share experiences is the more keenly felt (according to my empirical observations studying a sample group of four children) the more busy and distraced the intended recipient – the sharee, if you will – is at the time.
    Never mind how long it takes them to start talking. How do you get them to stop?

  31. my youngest sister used to talk like that when she was 5-6-7 yo, she would repeat the same story to all the family members even if they all were in one room and i used to fight with her all the time b/c am 8 yrs older than her, my older sister wouldn’t say anything, the younger one was treated by her as if they were the same age and i was the only authority for her cz our parents would allow her everything
    then she went one summer to the pioneer camp and returned all quiet, so i think time alone away from family helps and i got that time when i was 3 yo and was sent to my grandma’s place for the whole summer, returned talking her buriad dialect

  32. The most effective methods are illegal, Per. The heathen state has stepped between father and child, subverting God’s law and bringing forth countless generations of faithless miscreants. Get used to it.

  33. Re nurture vs Nature.
    Another part of my saga in surviving childhood.
    Nanny gave me milk but never said bah, but did butt in once in a while. Billy was a real stinker, no one would go near him except nanny but only when she was in the mood.
    But thanks to nanny I started to put some poundage.
    I being one that was lactose intolerant.
    But I was luckier than Remus, I never learnt to wolf it.
    We love to make rules that there be only one answer, impossible, they just be guidelines that maybe make over the 50 percentile just like marriage only a certain percentage are monogamist, rest have other agendas .
    Nature loves to have variation to find out what works..

  34. Angiportus says

    A lake talker? Is that something like a horse whisperer?

  35. michael farris says

    “we are talking about first language acquisition: persons who for one reason or another did not acquire language as small children have great difficulties not just with pronunciation but with grammar (word order, endings, etc).”
    I was kind of thinking out loud, wondering why the gap seems so much easier to close with sign. I do actually kind of think (with no evidence!!!) that motor skills (for lack of a better term) are part of it. I know that in learning a foreign language even when I intellectually know what to do my mouth and brain aren’t always on speaking terms (pun unintended) and what I actually say is often some distance away from what I’d intended (both in terms of word order and endings).
    Maybe there’s something about coordination of thought and mouth that has to be activated at a young age…. that doesn’t apply so much to the vicual gestural channel – IIRC the oddness of the langaugeless man’s ASL did have something to do with the finer grained aspects of non-manual features and expression of verb morphology (very much about subtle changes in length, direction and intensity of movement)

  36. To be fair, Einstein made an enormous and embarrassing mistake with the introduction of cosmological constant. So the late talking may have been a forewarning sign of this future blunder.

  37. I think too much is made of infancy milestones in general. My mother claims that by my first birthday I could say pretty much anything, but I certainly didn’t turn out to be anything special intellectually. My sons both spoke late (about two and a half before even simple sentences occurred) and they both ended up a good bit brighter than their mama.

  38. marie-lucie says

    It is usually assume that children have no knowledge of language until they start to utter what sounds to parents like an actual word, but they start understanding what they hear quite a bit earlier than they actually say anything, and until they achieve adult mastery their passive knowledge is much greater than their active ability. It is this accumulated passive knowledge that causes some late talkers to start uttering a perfectly formed sentence and then continue to talk.

  39. Paul Frank says

    Before she died, my mother told me that she could remember not talking when she was four and thinking that one of these days she was going to. And one day she did. And she went on to become a great story teller. The Einstein story is nice but probably too good to be true. Pretty much the same story is told about two-year-old who happens not be be Einstein in Andreas Hetzel, Negativität und Unbestimmtheit: Beiträge zu einer Philosophie des Nichtwissens (Transcript Verlang, 2009, p. 76).

  40. …accumulated passive knowledge…
    Nicely put! That accumulated passive knowledge is what I study. More specifically, I study children’s acquisition of the syntax of their first language.
    Since kids typically understand so much more than they can say, we usually use techniques that test children’s comprehension, rather than focusing on their production. There are so many other things that can get in the way with production when you’re under 3! Some of the sounds are still really hard to produce, your vocabulary’s bigger than it was, but it’s still pretty limited…

  41. …wondering why the gap seems so much easier to close with sign…
    Perhaps it’s not that sign-languages are easier to learn late, but that a larger percentage of the signing population came to the language late, and so it’s easier to blend in with the community as a late-learner.
    If you’re up for an academic article, you might take a look at the second one I linked for marie-lucie. It looks at native, early and late learners of ASL and shows that though they all get the basic sentence structure just fine, the late learners have significantly more difficulty with the verb morphology.
    (Though now that I’m trying to go back to it to double check that my description is right, I’m having trouble getting it to download. If other people want it, but can’t get at it, shoot me an email and I can find it elsewhere or send it to you directly.)

  42. I recently blogged a series of notes from two decades ago about our daughter’s language development from 8 months through 47 months. Up until about 18 months, she was mostly studying language, constantly eliciting it while communicating nonverbally. She did the same when learning a second language during a preschool year in Guangdong, China, when she was two. She was just on the verge of actually speaking some Mandarin and/or Cantonese but quickly abandoned that project as useless when we returned to the English-speaking world.
    Looking back through those notes, I’m struck by how slow her mastery of phonology (and language output more generally) was compared to other aspects of social communication, including reading and writing. (She’s still very bookish.) And how late (around 39 months) she started referring to herself as ‘I, me, my’ rather than Daytoh (Rachel).

  43. marie-lucie says

    CL: Since kids typically understand so much more than they can say, we usually use techniques that test children’s comprehension, rather than focusing on their production.
    Exactly, I am very glad you are doing this research! (Understanding much more than speaking is also typical of second-language learners).

  44. I highly recommend Joel’s series, which I’ve been following with great interest; here‘s the latest entry. (Joel, it would be useful if you’d provide an easy way to access them all, like a series link in your right column.)

  45. Trond Engen says

    Since kids typically understand so much more than they can say, we usually use techniques that test children’s comprehension, rather than focusing on their production.
    I think this is something every parent has done in some way, but more for play than for meticulous study. We used to give complicated messages like “Go into the kitchen and throw this napkin in the bin, the find a teaspoon in the second drawer and give it to granddad.” But we never took notes, so I don’t remember every test or exactly how old the children were at the time.

  46. The teacher duly arrived to test my grandson, with indeterminate results (she wasn’t very articulate in English herself, and had a thick probably-Slavic accent, according to the report I heard). I must add that Dorian has had a hearing test already, and he has normal hearing.
    As far as I can tell, he really doesn’t understand very much either: he can recognize specific sequences (“Elmo”; “Go to mommy!”; etc.), but hell, so can a trained dog. He doesn’t repeat what he hears, either, at least not with any more than chance likelihood.
    And for the parents with earaches out there, all my sympathies. I went through all that twenty years ago with Dorian’s mother, and I have two maxims to contribute: “For the first year you want them to walk and talk; after that, you want them to sit down and shut up!” and “Never have children, only grandchildren.”

  47. David Marjanović says

    To be fair, Einstein made an enormous and embarrassing mistake with the introduction of cosmological constant.

    Or so he thought.
    Turns out the expansion of the universe is accelerating.

  48. Dorian is six now, and his language is normal, or even a little past normal when it comes to his vocabulary. But under stress he still reverts to non-linguistic means, unfortunately including hitting. He has an ADHD diagnosis now, like his mother though not as severe, and medications are indicated.

Speak Your Mind