Firing and wiring
« previous post | next post »
In discussions about the history of usage, like this one, people often bring out generic memories ("I heard this all the time back in such-and-such a time period") or even more specific recollections ("I remember so-and-so saying this back in 19XX"). I've done this myself more than once. But recently something happened that made me wonder whether these memories can sometimes be false ones.
In reading a recent magazine story, I was taken aback by a parenthetical — Siddhartha Mukherjee, "Runs in the Family", The New Yorker 2/28/2016 (emphasis added):
The human eye is born restless. Neural connections between the eyes and the brain are formed long before a child is born, establishing the wiring and the circuitry that allow her to begin visualizing the world the minute she emerges from the womb. Long before the eyelids open, during the early development of the visual system, waves of spontaneous activity ripple from the retina to the brain, like dancers running through their motions before a performance. These waves reconfigure the wiring of the brain—rehearsing its future circuits, strengthening and loosening the connections between neurons. (The neurobiologist Carla Shatz, who discovered these waves of spontaneous activity, wrote, “Cells that fire together, wire together.”) This fetal warmup act is crucial to the performance of the visual system: the world has to be dreamed before it is seen.
This brought me up short because I associate the "what fires together, wires together" apothegm with Donald Hebb. So I checked the Wikipedia article on Hebbian theory, which confirmed the association but attributed the phrase neither to Hebb or to Shatz bur rather to a third person, Siegrid Löwel:
Hebbian theory is a theory in neuroscience that proposes an explanation for the adaptation of neurons in the brain during the learning process. It describes a basic mechanism for synaptic plasticity, where an increase in synaptic efficacy arises from the presynaptic cell's repeated and persistent stimulation of the postsynaptic cell. Introduced by Donald Hebb in his 1949 book The Organization of Behavior,[1] the theory is also called Hebb's rule, Hebb's postulate, and cell assembly theory. Hebb states it as follows:
Let us assume that the persistence or repetition of a reverberatory activity (or "trace") tends to induce lasting cellular changes that add to its stability.… When an axon of cell A is near enough to excite a cell B and repeatedly or persistently takes part in firing it, some growth process or metabolic change takes place in one or both cells such that A's efficiency, as one of the cells firing B, is increased.
The theory is often summarized by Siegrid Löwel's phrase: "Cells that fire together, wire together."
The citation is to Lowel, Siegrid, and Wolf Singer. "Selection of intrinsic horizontal connections in the visual cortex by correlated neuronal activity." Science 255, no. 5041 (1992): 209-212:
These results suggest that the development of tangential intracortical connections depends use-dependent selection mechanisms similar to those in the development of thalamocortical connections (16): neurons wire together if they fire together.
The Wikipedia article on Carla Shatz puts the pieces together:
Carla Shatz is credited with coining the sentence summarizing the Hebbian theory: "Cells that fire together, wire together." Although a similar phrase might first have appeared in print in Siegrid Löwel's Science article in January, 1992, Shatz had been using it in lectures for a number of years before. In her September 1992 Scientific American Article, she wrote: "Segregation to form the columns in the visual cortex […] proceeds when the two nerves are stimulated asynchronously. In a sense, then, cells that fire together wire together. The timing of action-potential activity is critical in determining which synaptic connections are strengthened and retained and which are weakened and eliminated"
So the story seems to be that Hebb invented the idea in the 1940s, and Shatz and/or Löwel coined the phrase around 1992.
My problem is that I distinctly recall learning about Hebbian learning theory from Eric Lenneberg in 1966 — and hearing him summarize the theory using the phrase "what fires together, wires together". I remember wondering whether this phrase had influenced or been influenced by "The family that prays together, stays together" (which apparently dates to about the same period in the 1940s as Hebb's work). And I remember thinking of the slogan again in the 1980s, when I learned about "back propagation" in connectionist networks, and wondered what if anything that process had to do with Hebb's "what fires together, wires together" idea.
But after spending a fruitless half hour in web search, I've convinced myself that these may be false memories.
It's certainly true that I took a course from Eric Lenneberg in 1966, and that he covered Hebb as well as other traditional learning theories. And it's certainly true that I learned about connectionist learning methods in the 1980s, and no doubt I wondered then about the connection to Hebb.
And surely the resonance between "prays together stays together" and "fires together wires together" is not an accident. So it makes sense that some version of "Cells that fire together, wire together" originated in the late 1940s or early 1950s, as an ironic Hebbian echo of the then-popular slogan "families that pray together, stay together"; that the Hebbian phrase circulated by word of mouth in a sort of psychological underground during a period when that sort of learning theory was out of fashion, but never made it into print; that I heard it from Lenneberg, who was a member of the community in which such a phrase might have circulated; and that the phrase emerged into public view in 1992, though being used by two different authors in a Scientific American article and in a paper in Science.
But there's no evidence that this story is true, or at least none that I could find. So alternatively, maybe the "fires together wires together" way of summarizing Hebb really does date to the early 1990s, and this evocative slogan somehow infiltrated back into my relevant memories from a quarter century earlier.
The truth of this particular history doesn't really matter much. But it's useful as an illustration of the uncertainty of linguistic memory.
Don Monroe said,
April 21, 2016 @ 7:18 pm
I had the same reaction to the statement when I read it in Mukherjee's generally fine story: that it was an old idea that I also associate with Hebb. Thanks for looking into it.
It may be relevant, though, that there are two somewhat distinct interpretations for "together." The Hebbian version refers to neurons that tend to have high levels of activity under similar circumstances. So "fire together" is statistical association of average firing rates.
There is more recent idea that neurons that are "bound" into some kind of coherent pattern tend to exhibit temporal correlations between their individual action potentials. In this case, "fire together" means that these individual events occur at about the same time. A particular stimulus could create high activity in two separate neurons without necessarily inducing tight correlations between their individual spikes.
I believe there is experimental support that both types of firing together can strengthen synaptic connections.
AntC said,
April 21, 2016 @ 7:22 pm
So the meme is: the NP that Vb1 together Vb2 together. That seems reasonably compositional.
It gains more force if the verbs are short and scan/rhyme, but is that essential?
Could it just be independent invention? Or perhaps in 1966 Lenneberg thought it too whimsical to put in a published paper? So it went underground for ~25 years.
Perhaps There's a LL commenter who also attended Lenneberg's lectures?
[(myl) I don't think that Lenneberg originated the phrase — if indeed I heard it from him, he was repeating a folkloric meme.]
Dick Margulis said,
April 21, 2016 @ 7:26 pm
A footnote on "The family that prays together stays together": The J. Walter Thompson ad agency was a member of the Ad Council, which distributed to its members requests for public service announcements. In about 1948, the call for a PSA for Religion in American Life went to JWT, and the assignment went to a creative group to come up with a slogan. A young copywriter by the name of Colin Campbell Dawkins (https://en.wikipedia.org/wiki/Colin_Dawkins)—who twenty years later hired me for my first job after college—was trying to come up with earnest inspirational slogans but not getting anywhere. Five o'clock was approaching, and in a cynical moment tossed off "The family that prays together stays together" as a sardonic joke that was sure to be rejected. It wasn't.
The Wikipedia article on Patrick Peyton implies that he coined the saying. He used it, but he didn't originate it. Dawkins did. According to the Wikpedia bio, Dawkins started in the mailroom after World War II and worked as a copyrwriter in the London office beginning in 1949. This is consistent with your estimate of the late 1940s for the slogan.
AntC said,
April 21, 2016 @ 7:31 pm
I should add: I'm rejecting as preposterous the hypothesis that Prof Liberman's memory is faulty ;-)
D.O. said,
April 21, 2016 @ 9:07 pm
It's not my cup of tea at all, but quick google books search shows that neurons were firing for a long time (since 1950s at least) they probably didn't wire until the meme, at least not very often.
ryanwc said,
April 22, 2016 @ 1:35 am
I'm trying to plumb this – so you're saying that you think the neurons in your brain to which the phrase "fires together wires together" map at some point fired together with other neurons that they had previously not fired together with, thereby creating a newly wired together (false) association for "fire together wires together"?
Bean said,
April 22, 2016 @ 6:59 am
So, has anyone done any studies on how long a word or phrase floats around in the vernacular (especially in some specific field) before anyone finally writes it down? There must be a typical timescale, especially "back in the day" when most written text was much more formal. Now we are much more likely to write things the way we would say them, and include slang/jargon (often explained) when it's the most compact way to express an idea.
Anyway, if that "spoken-to-written" timescale were on the order of decades then the whole story would hang together nicely.
I can think of many military terms and phrases that are essentially unGooglable, in that you can't find a definition for them, or the only definition is the one you wrote yourself in 2016 after talking to enough people yet they've been using the phrase for decades. And yet everyone in the military understands the phrase. Just because no one bothered to write phrases down (or make them publicly available!) doesn't mean they don't exist.
languagehat said,
April 22, 2016 @ 7:50 am
Good heavens, I've taken it for granted for years now that memories of such things can't be relied on; if there's one thing research into brains and memory have shown us, it's that we're constantly remaking our memories — they change every time we access them. I would certainly not accept any statement to the effect "I remember hearing this at X time," from myself or anyone else. (My wife and I have vivid, and contradictory, memories of where exactly in Grand Central Station we met for the first time.)
[(myl) We know that introspection is not a reliable guide to how people talk at the time of introspection, so it's hardly surprising that memory is not a reliable guide to what people heard or said decades ago. And episodic memory for exact wording is notoriously worse than memory for content.
But still, linguistic memory, like linguistic intuition, is not always completely wrong.]
Terry Hunt said,
April 22, 2016 @ 8:48 am
Ah yes, I remember it well.
Maurice Chevalier 1957
Murray Smith said,
April 22, 2016 @ 9:29 am
I am pretty sure I learned that phrase from Robert Hecht-Nielson in 1986, for what little that memory may be worth.
مدونة الارباح said,
April 22, 2016 @ 10:39 am
Besides the sheer injustice of it, acquiring a big pile of negative experiences in implicit memory banks, naturally makes a person more anxious, irritable and blue. Plus it makes it harder to be patient and giving toward others
bks said,
April 22, 2016 @ 10:41 am
Either I have false memories, or my wife does.
andyb said,
April 22, 2016 @ 11:30 am
@Bean: "So, has anyone done any studies on how long a word or phrase floats around in the vernacular (especially in some specific field) before anyone finally writes it down?"
But how would anyone study that? So little spoken text gets recorded; so little of that which is recorded gets kept around for decades; there's far fewer words for the same amount of storage (whether measured in physical size or in bytes once digitized); it's much harder to convert into something that can be indexed and searched; etc. So, there's no corpus to search the way there is for written text.
Bean said,
April 22, 2016 @ 11:57 am
@andyb: Well, yes I realize that it wouldn't be easy, but I'm not a linguist by profession, I have no idea what corpuses ARE available. Hence the question.
I bet individuals (collectively) might have quite a few random historical recordings… just think of all those people videotaping family events, graduations, hockey games, whatever, all through the 90s… there are soundtracks on all those videotapes and it's mostly unscripted. It would require an army of co-op students to transcribe, even if you could amass a decent sampling of recordings.
I personally had a couple of audio tapes from 1981-1983, in which I was recording random family scenarios (I was only in elementary school at the time). We've digitized what survived and it's really interesting to hear, (a) the change in my accent, (b) the change in my mother's accent, and (c) the unchanged level of sarcasm she used in interacting with us. :)
andyb said,
April 22, 2016 @ 1:43 pm
@Bean: The Internet Archive (the same people who do the Wayback Machine for dead websites) started a home movie archive. I think they've only got a few hundred digitized over the first decade, but they also have collections of orders of magnitude more movies sitting in film warehouses, some of them with grants allowing anyone interested to watch or digitize them, so with a big enough project, you could do this.
I assume we'll have a lot more for the 2020s and beyond. Technologically, the only thing stopping every phone call, video chat, CCTV feed, Nest-style home security cloud feed, etc. from being recorded is the cost of storage, which keeps coming down. Of course converting it all to text to be indexed and searched (or directly indexing audio speech) requires better algorithms and much more processing power, but those should be coming too. Socially, most people are reticent to allow governments and corporations to do that. But that won't matter in China. And in America, people will grumble about the Facebook-style privacy policies, but they'll keep using such sites, and upload more and more of their data. And even the ones who demand privacy will get used to having iCloud and similar sites hacked. And once it's possible to anonymize voice data the way as text and metadata, there will be even less resistance. So, in the future, if you want to know when everyone started using "vidchug" to mean playing a drinking game remotely over video chat synced to an interactive movie stream, you'll be able to find out that it was first mentioned in March 2023, exploded in popularity in September of that year among the frat-like online groups created by American students who can't afford to go to real college anymore and study at home for DeVryOnline certifications, and reached the general population in January 2024. :)
Hebbian learner said,
April 22, 2016 @ 2:21 pm
I never thought that Hebb coined the phrase because it isn't in his writings and it was not his style of writing or speaking. If Shatz coined or popularized the maxim, fine. However, the interpretation of the NYer text was ambiguous: it strongly invited the inference that she didn't just popularize a phrase, she discovered the concept to which it refers, which is untrue, because of the way it was juxtaposed with the discussion of her recent research. All due respect to Shatz, a superb scientist who I assume would not claim to have discovered Hebb's idea. Though Hayek did, as discussed prev. in this blog I believe. http://archive.boston.com/news/globe/ideas/articles/2004/01/11/friedrich_the_great/
D.O. said,
April 22, 2016 @ 3:46 pm
@bks: Both!
MonkeyBoy said,
May 1, 2016 @ 9:27 am
"The family that prays together, stays together"
That phrase pattern was ripe for reuse and spoofing. As a child I recall an early 1960s Mad Magazine parody of TV's Bonanza and its folksy wisdom that included "the family that bathes together stays together" along with a cartoon of the whole Ponderosa clan in a bathtub – bubbles and back brushes included. I seem to recall that parody had several other variations on the pattern. but that is the only one I recall.
A Google somewhat shows that "the family that bathes together stays together" has been taken up by nudism advocates.
Dick Margulis said,
May 1, 2016 @ 9:40 am
@MonkeyBoy: I recall several takeoffs during the sixties, notably "the family that sleeps together keeps together" and, in the wake of the Sharon Tate murder, "the family that slays together stays together." As for Mad Magazine, Colin Dawkins and Harvey Kurtzman were friends, and Dawkins had a sense of humor regarding parodies of the slogan. It wasn't something he was particularly proud of; he was always bemused by its having been selected in the first place.