Hess and Markram didn't invent topological analysis of neuronal networks. They are not mapping the brain. This is a fluff piece that didn't bother to check with anyone else in the field if this paper is as groundbreaking as it's being presented, which is the absolute minimum of journalistic rigor.
The claim this ended up in Frontiers because no one else in neuro theory has heard of algebraic topology is laughable.
e: To be clear, I don't think this rises to the level of "Sokal hoax" -- it looks like an interesting result. I also have an interest in the role massive simulation can play in discovery. But to misrepresent topological analysis as an entirely novel approach or package it with phys.org-press-release nonsense like "mapping the brain in 11 dimensions" is both a disservice to the rich field on which it's building and adds noise to public communication about neuroscience, which is already plagued with bullshit.
The unfortunate pattern in pop science is the hero complex: the journalist learns about a field and feels a need to give all the credit to one person or small team, and to five them an adversary, to make the story seem dramatic, revolutionary, yet accessible. The real stories are too large and complex to make exciting in a quick read to a layperson.
Hmmmm. "Hess says they tried to shop around the idea to other neuroscience journals first, to avoid the conflict of interest—but didn’t have much luck, since most editors they spoke to had never heard of algebraic topology".
Neuroscience as a field is quite comfortable with 'n-dimensions', given the importance of matrices to data. Let alone the relationship between topology and graph/network theory... It seems highly unlikely that editors would not be at least roughly familiar with algebraic topology. It reads (to me), "Our research is not marketable, because it lacks prescriptive insight, so we published it in our own journal". The conflict of interest statement in the paper is suspect, as the article roughly mentions.
From the abstract: "We propose that the brain processes stimuli by forming increasingly complex functional cliques and cavities". I can't see how this says anything not already well-established, though perhaps differently described - that biological structures in the brain support the use of hierarchical network models at the neuronal IP level.
Anecdotally, I find the 'jargonification' of neuroscience a little disturbing. This reads like false progress. I'm simply an interested lay-person, so I'm open to correction on all the above - but as I see it, additional ways of describing 'what' does not necessarily give a 'how'.
> I can't see how this says anything not already well-established
I am not a neuroscientist. It sounds like the cavities are orderly in 11 dimensions, whereas previously they seemed chaotic - the "shape of the activity" (i.e. something akin to a fMRI) is structured. This is different to the mathematical model of a neuron (being the matrix you mentioned).
This article about mapping the brain used the expression 'lol' in the first paragraph. I don't think it's nitpicking to say that significantly undercuts the seriousness with which I read the article.
The lol is perfectly appropriate in the context it is used:
KATHRYN HESS CAN’T tell the difference between a coffee mug and a bagel. That’s the old joke anyway. Hess, a researcher at the Swiss Federal Institute of Technology, is one of the world’s leading thinkers in the field of algebraic topology—in super simplified terms, the mathematics of rubbery shapes. It uses algebra to attack the following question: If given two geometric objects, can you deform one to another without making any cuts? The answer, when it comes to bagels and coffee mugs, is yes, yes you can. (They only have one hole apiece, lol.)
I'm not sure what the people who are taken aback by it usually read -- or what did they expect to find here.
> I'm not sure what the people who are taken aback by it usually read -- or what did they expect to find here.
They don't understand what they're reading, so instead of asking for an explanation, they attack the lowest hanging fruit (the 'lol').
It's just a form of position signalling, basically they assume everyone else understands it and they want to be in that 'in-group', so they try and look smart by tearing it down by attacking nothings.
That 'lol' stopped me in my tracks and I thought seriously about commenting but decided (incorrectly) it was too trivial a nitpick. I'm glad you spoke up.
The other thing I thought about the article is that I hate being told what the Sokal Hoax is _every time it's mentioned_, doesn't everyone know by now? Also, aren't most people (okay, the sort of person who's reading a Wired article on computational neuroscience) going to know that bagels and coffee-cups are topologically identical? It's not that esoteric is it? And that's happening more and more these days. I think if you're writing about complex neuroscience you've got to assume your audience knows about these things. Is that a small or big gripe, I'm not sure. Grumblegrumble.
I, for one, did not know what any of these things were prior to reading the article. I think Wired is focused at an "interested but not necessarily experienced" audience, so I appreciate the clarifications.
I'd argue against the use of LOL for any writing that involves an exchange of money, at any level, reader, author, publisher, printer, paywall, et c.
But with regard to trivia, such as The Sokal Hoax, it is always, always, always the job of the journalist to inform the reader. This is true for acronyms, units of measurement, coincidental circumstances and more.
If someone writes an article about the JFK assassination, they should be explaining JFK's initials, events surrounding the cuban missile crisis, the MLK assassination, Jack Ruby's mob ties, LBJ's rivalry, Bobby Kennedy's assassination, Ted Kennedy's car accident, the grains of the bullet used, what the unit of measurement known as the grain measures for a bullet, even what a knoll is.
The journalist should capture as much information as needed to convey all relevant information to tell a complete story. A bad journalist leave details out, and expects the reader to "just know everything."
Agreed on LOL. However, just to play devil's advocate – algebraic topology was directly related to the story so I guess it needed explaining, whereas the Sokal Hoax was invoked by a correspondent
> Chris Eliasmith, a theoretical neuroscientist at the University of Waterloo, didn’t find much to be interested about in his first reading: “It strikes me as an over-stated and minor contribution,” he wrote in an email. Peter Latham, a neuroscientist at the University College London went one step further, saying that it “reads a bit like an Alan Sokal hoax”
Furthermore, the explanation —referring to the NYU mathematical physicist who got a bogus article about quantum physics as a "social construct" published in an academic journal — doesn't convey the salient details about the Sokal Hoax (the style of language used, the reliance on science analogies to convince a non-science audience, the purpose of the hoax (to make a point about what they saw as a certain type of bullshit article, and so on).
But I totally take your point – and together with the other reply I stand corrected.
Only distantly related, but the title reminds me of line from AI (pseudo-?) researcher Chris McKinstry[1]'s web suicide note:
>Oh and BTW, the mind is a maximum hypersurface and thought a trajectory on it and the amygdala and hippocampus are Hopf maps of it. No one knew this before me, and it seems no one care. So be it. My time will come in a hundred or a thousand years when the idea again returns.
It is waste of time to build telescope. What is the point. See the ship up there. I know it is there.
What is the point of big data. What is the point of neural network. Waste of time.
Until some point the telescope to what it is forbidden and said the forbidden by a guy called ... what are those things around Jupiter. It can't exist as it is not revolve around the earth
What is go? Silly game?
What is car drinking we can do it
... I guess to this silly scientific community it is just do not see beyond the present paradigm.
This may provide a way to understand the upcoming AI via ... understanding. What a waste of time!
The claim this ended up in Frontiers because no one else in neuro theory has heard of algebraic topology is laughable.
e: To be clear, I don't think this rises to the level of "Sokal hoax" -- it looks like an interesting result. I also have an interest in the role massive simulation can play in discovery. But to misrepresent topological analysis as an entirely novel approach or package it with phys.org-press-release nonsense like "mapping the brain in 11 dimensions" is both a disservice to the rich field on which it's building and adds noise to public communication about neuroscience, which is already plagued with bullshit.