The Penicillin Myth
asimov.press

・

100 points

・

surprisetalk

・

5 hours ago

51 comments

glenstein ・ an hour ago

In my opinion, we're possessed by a cultural epidemic of think pieces doing rich and nuanced science history, but wrongly framed in the form of correcting "myths" that, in their substance amount to quibblings over narrative emphasis. It's easy to get taken in by the framing because it truly is enlightening, and the argument goes down so smooth because its embedded in a rich, curious, and fascinating scientific history that otherwise embodies best practices I would happily celebrate.

But the key details about the story of penicillin are that a moldy plate showed bacteria-free clearing, Fleming saw it, isolated the mold, proved its germ-killing filtrate and published the finding, which is the heart of the story and which is not a myth.

I'm sure it's true enough that St Mary's windows were usually kept shut to keep pathogens in and contaminants out, that London's August 1928 cold snap would have slowed staph growth, that Fleming's first notes Or 8 weeks later than the actual event, and that a modern plate seeded with bacteria first will not produce the celebrated halo unless the mold is given a head start. The article makes much of the fact that today’s researchers cannot reproduce the famous halo if they add staph first, yet that difficulty rebuts a sequence Fleming never claimed to have used.

These points are significant, even fascinating, yet the article inflates them into a strobe-lit "MYTH" banner, turning normal human imprecision about times and temperatures into evidence of wholesale fiction, which abuses the ordinary friction of any retrospective account and punishes the story for the very human messiness that makes it instructive.

The window quibble, the incubator gap, and the replication protocol do not touch the central, uncontested fact that chance contamination plus observational curiosity gave medicine its first antibiotic.

jjk166 ・ an hour ago

Hare's theory predicts that there would need to be a cold snap at just the right time, and lo and behold there was. Probability isn't an issue if the only reason you are considering the probability is because the event already happened. Indeed the low probability of such an event transpiring goes a long way towards explaining why the discovery was not made earlier.

Root-Bernstein's theory makes no such testable predictions, and it solves the issue of an incomplete record on September 3rd with incomplete or inaccurate records elsewhere. It seems to me extremely plausible that fleming did not record the results of a botched, uncontrolled experiment but still recognized it as an indicator of something interesting that warranted follow-up. If I were in his position I would preserve the random dish for comparison to the more rigorous follow up experiment. I certainly don't put any stock into the argument that if the story had gone as Root-Bernstein describes it would have been too circuitous for scientific publishing, if anything it would be much more harmonious with standard scientific writing than the chance observation story.

arjie ・ 4 hours ago

Oh I really enjoyed this one.

Got a quick insight about how penicillin works: interferes with cell-wall building which is a destroy and recreate process by preventing the recreate part.

Got a quick view into the scientific process and communication: Fleming focused on the insight - penicillium kills staphylococcus - and left out the circuitous detail. This is important so that the big win here is very clear.

And got an insight into human nature and memory: Fleming didn’t tell the accidental contamination story until much later. It could possibly be even an idea someone else might have come up with which then took root in his mind (ironic haha!)

The communication aspect reminds me of Mendel’s far too perfect ratios for his pea plants. That kind of “repeat till difference clear” statistics would be decried today but perhaps that was to communicate rather than to determine.

And finally, I really enjoy reading about human process innovation because I think it’s a big factor in how Humanity grows. The lab notebook has to be some kind of star performer here - Fleming’s notes allow us to look back like this.

When I experiment with things, I naturally lean to keeping notes on my test protocol, observations, and results. But not because of some personal genius. It’s just the standard way I was taught as a child in our science labs.

I won’t claim to the rigor of a microbiology lab but even just the process notes help a lot, which is useful since I’m just testing molecules on myself.

  • stevenwoo ・ an hour ago

    If you are not familiar with more of Mendel or plant biology, he got extremely lucky in picking a two chromosome species. The next plant he picked had more than two chromosome types so he spent the rest of his life hitting his head against the wall - obvious to us but him not having a theory and expertise with microscopes to explain his pea results hampered him greatly beyond his initial pea plant studies.

  • awkward ・ an hour ago

    So many other good details that get to how impossibly multivariate biology research is, like the need to have several days at the exact temperature.

    It's not uncommon for results in biology to have this kind of snag in reproducibility even now. Sometimes it's due to attributing variations to something like "steady hands at the bench", but other times it can even be a deliberate attempt to prevent rivals from duplicating a process before it can be patented and privatized.

  • Joker_vD ・ 4 hours ago

    > Mendel’s far too perfect ratios for his pea plants.

    "Remember, if you flip a coin 200 times and it comes heads up exactly 100 times, the chances are the coin is actually unfair. You should expect to see something like 93 or 107 instead".

    • fainpul ・ 3 hours ago

      Isn't 100 / 100 the most likely outcome for a fair coin? Sure it's unlikely that you hit exactly that result, but every single other result is even less likely.

      What I'm trying to say: if you get 100 / 100, that's not a sign of an unfair coin, it's the strongest sign for a fair coin you can get.

      • shakow ・ 3 hours ago
        ・ 11 more

        > every single other result is even less likely.

        But the summed probability of the “not too far away results” is much higher, i.e. P([93, 107]\{100}) > P([100]).

        So if you only shoot 100/100 with your coin, that's definitely weird.

        • tshaddox ・ 2 hours ago
          ・ 2 more

          Okay, but it doesn't make sense to arbitrarily group together some results and compare the probability of getting any 1 result in that group to getting 1 particular result outside of that group.

          You could just as easily say "you should be suspicious if you flip a coin 200 times and get exactly 93 heads, because it's far more likely to get between 99 and 187 heads."

          • wat10000 ・ 2 hours ago

            It's suspicious when it lands on something that people might be biased towards.

            For example, you take the top five cards, and you get a royal flush of diamonds in ascending order. In theory, this sequence is no more or less probable than any other sequence being taken from a randomly shuffled deck. But given that this sequence has special significance to people, there's a very good reason to think that this indicates that the deck is not randomly shuffled.

            In theory terms, you can't just look at the probability of getting this result from a fair coin (or deck or whatever). You have to look at that probability, and the probability that the coin (deck etc.) is biased, and that a biased coin would produce the outcome you got.

            If you flip a coin that feels and appears perfectly ordinary and you get exactly 100 heads and 100 tails, you should still be pretty confident that it's unbiased. If you ask somebody else to flip a coin 200 times, and you can't actually see them, and you know they're lazy, and they come back and report exactly 100/100, that's a good indicator they didn't do the flips.

        • fainpul ・ 3 hours ago
          ・ 8 more

          > So if you only shoot 100/100 with your coin, that's definitely weird.

          Not if you only try once.

          • shakow ・ 3 hours ago
            ・ 6 more

            Even if you shoot only once, you still have a higher chance of hitting something slightly off the middle than the perfect 100/100. And this because that's one point-precise result (100/100) vs. a cumulated range of individually less-probable results, but more probable when taken as a whole.

            For a fair coin, hitting 100/100 is ~5%, vs. ~30% falling in [97; 103] \ {100}. You can simulate here: https://www.omnicalculator.com/statistics/coin-flip-probabil...

            • tshaddox ・ 2 hours ago
              ・ 4 more

              > you still have a higher chance of hitting something slightly off the middle than the perfect 100/100

              That's because "something slightly off the middle" is a large group of possible results. Of course you can assemble a group of possible results that has a higher likelihood than a single result (even the most likely single result!). But you could make the same argument for any single result, including one of the results in your "slightly off the middle" group. Did you get 97 heads? Well you'd have a higher likelihood of getting between 98 and 103 heads. In fact, for any result you get, it would have been more likely to get some other result! :D

              • zdragnar ・ 2 hours ago
                ・ 3 more

                > But you could make the same argument for any single result

                Isn't that the point? The odds of getting the "most likely result" are lower than the odds of getting not the most likely result. Therefore, getting exactly 100/100 heads and tails would be unlikely!

                • tshaddox ・ an hour ago

                  But as I said, getting any one specific result is less likely than getting another other possible result. And the disparity in likelihoods is greater for any one specific result other than the 50% split.

                • alwa ・ 27 minutes ago

                  I think the disagreement is about what that unlikeliness implies. "Aha! You got any result? Clearly you're lying!"... I'm not sure how far that gets you.

                  There's probably a dorm-quality insight there about the supreme unlikeliness of being, though: out of all the possible universes, this one, etc...

            • e12e ・ 35 minutes ago

              Should that be 25% for 97..193 excluding 100?

          • kalaksi ・ 3 hours ago

            I'm sorry, but try what once? 200 flips once?

      • voakbasda ・ 3 hours ago

        I would guess that a single trial of 200 flips can be treated as one event, so getting 100/100 is but one outcome. It may be the most likely individual outcome, but the odds of getting that exact result feel less likely than all of the other possible outcomes. The 100/100 case should be seen the most over repeated trials, but only marginally over other nearby results.

        Intuitively, this seems right to me, but sometime statistics do not follow intuition.

      • dooglius ・ 3 hours ago
        ・ 4 more

        "fair coin" refers to both the probability of heads and tails being equal (which is still justified) as well as the trials being independent (unlikely with 100/200; more likely the "coin" is some imperfect PRNG in a loop)

        • tshaddox ・ 2 hours ago
          ・ 3 more

          > more likely the "coin" is some imperfect PRNG in a loop

          "More likely"? How can you even estimate the likelihood of the coin being "an imperfect PRNG" based on a single trial of 200 flips?

          • dooglius ・ 2 hours ago

            You can use combinatorics to calculate the likelihood. If your PRNG is in a cycle of length N in its state space (assuming N>200), and half the state space corresponds to heads (vs tails), then the likelihood would be (N/2 choose 100)^2/(N choose 200) versus your baseline likelihood (for a truly random coin) of (200 choose 100)/2^200.

            Graphing here https://www.wolframalpha.com/input?i=graph+%28%28N%2F2+choos... and it does look like it's only a slight improvement in likelihood, so I did overstate the claim. A more interesting case would be to look at some self-correcting physical process.

          • ajuc ・ 2 hours ago

            Bayesian vs frequentist in a nutshell :)

            If a student was tasked with determining some physical constant with an experiment and they got it exactly right to 20th decimal place - I'll check their data twice or thrice. Just saying. You continue believing it was the most likely value ;)

      • Joker_vD ・ 3 hours ago
        ・ 6 more

        Well, yes. But the expected deviation from the mean is still ≈7.07. And the probability that the outcome will be either 93/107 or 107/93 is (slightly) higher than the outcome being exactly 100.

        • fainpul ・ 3 hours ago
          ・ 4 more

          But those are 2 results. 100 / 100 is more likely than 93 / 107 (or any other specific result) is what I'm saying.

          • kalaksi ・ 2 hours ago

            I don't think that makes 100 / 100 the most likely result if you flip a coin 200 times. It's not about 100 / 100 vs. another single possible result. It's about 100 / 100 vs. NOT 100 / 100, which includes all other possible results other than 100 / 100.

          • paddleon ・ 3 hours ago

            You can tell the difference between

            93 heads, 107 tails and

            93 tails, 107 heads

            but not between

            100 heads, 100 tails and

            100 tails, 100 heads.

            So while those are two results (93/107 and 107/93), they really only count as two separate outcomes if you pre-specify that the first number is heads.

            If instead you consider symmetries, where there are 2 ways to get 93/107 and only one wall to get 100/100, then you have more likelihood for the 93/107 outcome because you have two ways to get it.

          • Joker_vD ・ 3 hours ago

            Depends what you count as a result, I guess. "There is exactly N flips of a single kind" is also a viable definition, just as "The exact sequence of flips was x_0, x_1, ... x_199" is.

        • eviks ・ 3 hours ago

          Why not go one abstraction further and go expected deviation from deviation? Probably the word "expected" plays a mind trick? "Expected" doesn't mean the probability increases, the easiest way to understand it is just by looking at the probability distribution function chart for coin tosses - you'll immediately see that mean has the highest chance of happenning, so exactly 100 is the most likely outcome

    • munchbunny ・ 3 hours ago

      The chance of exactly 100 heads from 200 fair coin flips is approximately 5-6%. Qualitatively, that's not particularly strong evidence for an unfair coin if you did only one trial.

      You could also argue that 100 out of 200 on a fair coin is more likely than any other specific outcome, such as 93/200, so if the argument is that the coin is "too perfect", you then also have to consider the possibility that the coin is somehow biased to produce 93/200 much more often than anything else, vs. 100/200.

    • n1b0m ・ 3 hours ago

      In a real-world scenario, if you saw a result significantly far from 100 (like 150 heads), you might suspect the coin is unfair. However, seeing exactly 100 heads gives no reason to suspect the coin is unfair; it's the result most consistent with a fair coin.

alexpotato ・ 2 hours ago

One of my favorite "myths" about the discovery of stainless steel:

Metallurgist is trying out all kinds of steels looking for a particular attribute. He would dutifully record each recipe + test in a notebook but if a particular batch didn't have the attribute, he would throw it out a window into an outdoor scrap pile.

Several months go by and he's cleaning up the pile and notices that one of the blocks has no rust or corrosion. He knows that the pile is six months old but doesn't know which of the recipes this block was connected to.

So he repeats ALL of the block recipes from the last 6 months but labels each block so he can figure out which recipe led to the "stainless" steel.

(Probably not the real story but always loved this telling of it. Actual Wikipedia history is here: https://en.wikipedia.org/wiki/Stainless_steel#History)

  • commandlinefan ・ 30 minutes ago

    I recall reading that the microwave oven was invented by a physicist after he walked by a radiation chamber and the chocolate bar in his pocket melted... makes me wonder if there was any historic license taken in that case as well.

mandevil ・ an hour ago

Best take on serendipty versus effort in the history of science is, naturally enough, from Monty Python:

https://www.youtube.com/shorts/1Hu0f_ti9EQ

Text from http://montypython.50webs.com/scripts/Series_3/99.htm

"Presenter[John Cleese]: Penguins, yes, penguins. What relevance do penguins have to the furtherance of medical science? Well, strangely enough quite a lot, a major breakthrough, maybe. It was from such an unlikely beginning as an unwanted fungus accidentally growing on a sterile plate that Sir Alexander Fleming gave the world penicillin. James Watt watched an ordinary household kettle boiling and conceived the potentiality of steam power. Would Albert Einstein ever have hit upon the theory of relativity if he hadn't been clever? All these tremendous leaps forward have been taken in the dark. Would Rutherford ever have split the atom if he hadn't tried? Could Marconi have invented the radio if he hadn't by pure chance spent years working at the problem? Are these amazing breakthroughs ever achieved except by years and years of unremitting study? Of course not. What I said earlier about accidental discoveries must have been wrong. "

mordae ・ 4 hours ago

Hmm, sounds like the spores must have come from the outside. Otherwise he'd be saying his colleague has contaminated the building with improperly stored fungal colonies and he himself let those spores contaminate his lab. So yeah, definitely from the outside.

keepamovin ・ 2 hours ago

Tangentially related: doxycycline helps improve muscle and tendon tear recovery by enhancing the performance of matrix proteins that form "scaffolding", by inhibiting factors that break them down. "By inhibiting MMPs, doxycycline helps preserve and remodel the ECM, accelerating repair and improving biomechanical strength (e.g., tensile strength and reduced creep/strain)". Crazy. I'm a big fan off high-ROI off-label uses of well-tolerated, cheap, out-of-patent "WHO essential medecines list" pharmas. There's much to be found there.

Another tidbit: inderol (propranolol, beta-blocker) can aide PTSD recovery by reducing the emotional potency of traumatic memories when taken in a therapeutic replay.

mandevil ・ an hour ago

The cartoon 60% of the way down the article definitely feels like an "artists barely concealed fetish" thing.

dj_gitmo ・ 4 hours ago

The Hare theory is a better story, regardless of whether it is true. I am surprised that it hasn't seeped into pop science lore.

niemandhier ・ 3 hours ago

This stood out to me:

“Despite this close professional association, however, Hare claims to have played no part in the discovery or original research on penicillin nor to have discussed them with Fleming”

It’s nice to see that the bickering about who stole whose research does not affect all old discoveries.

  • dooglius ・ 3 hours ago

    Is it common in cases for someone with no involvement at all to claim involvement? Usually disputes I've heard of are when multiple people are involved, and they're arguing about who played a crucial vs minor role.

    • jjk166 ・ 2 hours ago

      It's common for people who have close associations to have events that could be construed as involvement, and when someone does believe they are involved in something important, they tend to claim their involvement was important. It would be so easy to inflate a random conversation or a little common courtesy assistance as something more. It takes some genuine humility to take stock of all your interactions and conclude that you had nothing to do with one of the most important discoveries in history, and more still to admit that you thought nothing of it at the time.

johnyzee ・ 2 hours ago

That was a lot of words to get to the point that Fleming probably misremembered the sequence of events when he retold the story 15 years later. He even mentioned this possibility at the time. Interesting article but not much of a mystery.

tetris11 ・ 4 hours ago

(Off-topic:) Scrollbars, and their non-existence.

  • jasonjmcghee ・ 4 hours ago

    It's not the best implementation, but it's there (you just have to scroll a bit to get it to show).

dooglius ・ 3 hours ago

Can someone with more experience in scientific writing comment on

> It’s too circuitous and indirect for a scientific report

The preceding paragraph does not seem unreasonable to me--maybe a bit too glib, but nothing that couldn't be touched up.

  • paddleon ・ 3 hours ago

    in scientific writing, you need to take

    100 articles at 6 pages each

    and condense them to 1 article of 6 pages

    And add how your method/insight moves the conversation forward, along with describing your method/insight.

    The reason why scientific writing can be hard to read from the outside is the 100:1 compression. Decompression of that can require some knowledge of the field.

    Also, some people are just bad at writing.

  • advisedwang ・ 2 hours ago

    The goal of a paper is to write up the actual discovery, not to tell a story or explain irrelevant background. The steps he wrote there would confuse rather than elucidate the actual discovery.

tolerance ・ 3 hours ago

Writing like this does not belong on Substack.