Related to the UK post office scandal. It has been said that the managers at the post office long suspected that postmasters were cheating. Why else would anyone do such a demanding job for so little money? Surely not just as a service to their community? So basically projecting their own greed and morals on to the sub post masters. When 'evidence' from the software suddenly revealed thousands of them were actually cheating, it just confirmed what the managers already believed. They were then able to bring their own prosecutions, regardless of the flimsiness of the evidence.
I wonder if the postmaster prosecution case was influenced by the large number of supposed participants. If the bug had identified only a handful of "fraudsters" it would have just been a minor fluke in an organization and handled appropriately. But with thousands of bad actors, it borders on a grand criminal conspiracy. So there was an incentive for each individual prosecutor to aggressively pursue the cases under their jurisdiction so that nobody could lay blame for the whole on their respective part not being handled.
Yeah, perhaps after a while the earlier convictions gave "momentum" to later ones. It's human nature to think that after a ton of people have been convicted on the basis of some evidence that that evidence must be valid.
> It’s actually pretty easy for prosecutors to convict an innocent person if they really want to, as long as they have some kind of vaguely-incriminating evidence.
This theory predicts that trials almost always lead to guilty verdicts. But this seems false. The clearest source I can find says guilty verdicts occur in only 71% of magistrate court trials and 56% in Crown Courts.
I'm not sure about "almost always" but still—good point. Those stats appear to be pretty specific (for theft charges that could go to magistrate or crown courts) and I can't easily find more general stats aside from the fact that ~85% of people who are charged settle their cases. Still, I think you're right that this is is some evidence against that theory.
Type species exist because the concept of species itself is problematic. Species attempts to impose hard boundaries on a fuzzy biological continuum. The pyramid resting on top of this shaky foundation is equally shaky.
A tree is the simply the wrong data structure. But we've been using it for centuries so we need weird hacks like type species to keep the legacy codebase running.
Well there definitely some large differences that makes a classification warranted. But the problem is that since you cannot really know the boundaries because it would require studying basically every specimen, you have to create somewhat arbitrary boundaries.
Thanks for sharing this. The weird jargon and circular explanations of some wikipedia pages is really frustrating. I can’t really understand who they’re writing for. I understand why this type of obfuscation is in scholarly journals (though it’s probably still unnecessary), but not in what is essentially a reference source for non-specialists and laypeople. I’ve been using Chat-GPT to un-jargon and de-pretentious wikipedia pages. Simple Wikipedia and this extension are a great alternative.
I think this shows pretty well why the original is written the way it is. The thing that makes the simple better isnt the simpler words, its that it explains more of how the nomenclature works in general. The point of a hyperlinked encyclopedia is not repeating that stuff everywhere its relevant.
...that said, the dedicated page there is not much better.
These old visualizations are so cool. I personally believe that the explainer post will never die because it would and has somehow never occurred to me to investigate how Michaelangelo or other sculptors produce flawless work with no chips or imperfections!
I'm not ashamed to tell you that I'm not 100% sure. But as far as I can tell: Biologists have decided that each genus should have a single species called a "type species". When deciding if another species should be considered in the genus or not, you're supposed to compare it to the type species. If similar—good candidate for being in the genus.
Perhaps the British post office/government employees were not aware of the concept of bugs and attributed near-infallibility to the system? I could imagine a tech-illiterate org doing that in 1999.
Yeah, as far as I can tell they genuinely believed the system was 100% correct. Still, it's hard for me to understand how they convinced so many judges, how it took so long for there to be a media backlash, etc.
Apparently there's a UK legal presumption that computer evidence is reliable. (As of Jan 2025, this is under review.) It's also possible some Fujitsu expert witnesses may eventually be charged with perjury. The analogy of a DNA testing lab giving false positives seems a fair one.
While nobody "wanted" to hurt the postmasters, plenty had strong incentives to prefer an individual conviction to an admission that their expensive IT system was fundamentally unreliable, or (eventually) that they had been responsible for sending innocent people to prison (or worse).
Finally, the UK justice system is indeed very slow at fixing mistakes when they do happen. Other recent examples: Andrew Malkinson (after 17 years), Tom Hayes (10 years), Peter Sullivan (38 years), maybe Lucy Letby.
I picked it up from your quoted excerpt, but I was already familiar with the general concept from outside biology. That said, I might be misunderstanding it entirely and just didn't realize the excerpt wasn't what I expected it to be :D An example like "the grey wolf (canis lupus) will always be the species to which relative genetic distance determined membership in the family."
everyone please stop talking about the definition of type species i beg you
Related to the UK post office scandal. It has been said that the managers at the post office long suspected that postmasters were cheating. Why else would anyone do such a demanding job for so little money? Surely not just as a service to their community? So basically projecting their own greed and morals on to the sub post masters. When 'evidence' from the software suddenly revealed thousands of them were actually cheating, it just confirmed what the managers already believed. They were then able to bring their own prosecutions, regardless of the flimsiness of the evidence.
I wonder if the postmaster prosecution case was influenced by the large number of supposed participants. If the bug had identified only a handful of "fraudsters" it would have just been a minor fluke in an organization and handled appropriately. But with thousands of bad actors, it borders on a grand criminal conspiracy. So there was an incentive for each individual prosecutor to aggressively pursue the cases under their jurisdiction so that nobody could lay blame for the whole on their respective part not being handled.
Yeah, perhaps after a while the earlier convictions gave "momentum" to later ones. It's human nature to think that after a ton of people have been convicted on the basis of some evidence that that evidence must be valid.
> It’s actually pretty easy for prosecutors to convict an innocent person if they really want to, as long as they have some kind of vaguely-incriminating evidence.
This theory predicts that trials almost always lead to guilty verdicts. But this seems false. The clearest source I can find says guilty verdicts occur in only 71% of magistrate court trials and 56% in Crown Courts.
https://www.adruk.org/news-publications/publications-reports/data-insight-deciding-to-have-a-crown-court-jury-trial-for-a-theft-offence-consequences-and-relationships-with-ethnicity/#:~:text=Figures%20from%20the%20Crown%20Prosecution,heard%20in%20the%20Crown%20Court.
I'm not sure about "almost always" but still—good point. Those stats appear to be pretty specific (for theft charges that could go to magistrate or crown courts) and I can't easily find more general stats aside from the fact that ~85% of people who are charged settle their cases. Still, I think you're right that this is is some evidence against that theory.
Type species = trope namer? https://tvtropes.org/pmwiki/pmwiki.php/Main/TropeNamers
I think that's a good analogy yes.
Type species exist because the concept of species itself is problematic. Species attempts to impose hard boundaries on a fuzzy biological continuum. The pyramid resting on top of this shaky foundation is equally shaky.
A tree is the simply the wrong data structure. But we've been using it for centuries so we need weird hacks like type species to keep the legacy codebase running.
Well there definitely some large differences that makes a classification warranted. But the problem is that since you cannot really know the boundaries because it would require studying basically every specimen, you have to create somewhat arbitrary boundaries.
type specimen : species :: type species : genus
Thing explainer posts will never die, because it's some random dude explaining shit how they see it, not some omniscient god-jerk (AI).
Good thing there's a Simple English Wikipedia explanation for Type Species: https://simple.wikipedia.org/wiki/Type_species
I wrote a Chrome extension to let me know if a Simple version of an article exists: https://mattsayar.com/simple-wikiclaudia/
Thanks for sharing this. The weird jargon and circular explanations of some wikipedia pages is really frustrating. I can’t really understand who they’re writing for. I understand why this type of obfuscation is in scholarly journals (though it’s probably still unnecessary), but not in what is essentially a reference source for non-specialists and laypeople. I’ve been using Chat-GPT to un-jargon and de-pretentious wikipedia pages. Simple Wikipedia and this extension are a great alternative.
This is definitely better. It's funny the the idea of a type specimen is so central there, when it barely appears in the main article.
It's also funny that while the simpler article links to https://simple.wikipedia.org/wiki/Type_specimen, the main article separately links to https://en.wikipedia.org/wiki/Type_(biology) and https://en.wiktionary.org/wiki/specimen !
I think this shows pretty well why the original is written the way it is. The thing that makes the simple better isnt the simpler words, its that it explains more of how the nomenclature works in general. The point of a hyperlinked encyclopedia is not repeating that stuff everywhere its relevant.
...that said, the dedicated page there is not much better.
These old visualizations are so cool. I personally believe that the explainer post will never die because it would and has somehow never occurred to me to investigate how Michaelangelo or other sculptors produce flawless work with no chips or imperfections!
This is high quality retro internet usage. Thank you!
Ooh, surprise h/t, thanks!
(minor correction: Premarin, not Pregmarin)
Thank you, fixed!
So wait, what *is* a type species? I thought you were going to explain it plain language so we could see how it should be done.
I'm not ashamed to tell you that I'm not 100% sure. But as far as I can tell: Biologists have decided that each genus should have a single species called a "type species". When deciding if another species should be considered in the genus or not, you're supposed to compare it to the type species. If similar—good candidate for being in the genus.
Sounds like the medoid, except the geometry of the space is only vaguely hinted at.
Yes, except I think the "causality" is sort of the opposite: A set defines a mediod, whereas a genus is (sort of) defined by the type species.
Perhaps the British post office/government employees were not aware of the concept of bugs and attributed near-infallibility to the system? I could imagine a tech-illiterate org doing that in 1999.
Yeah, as far as I can tell they genuinely believed the system was 100% correct. Still, it's hard for me to understand how they convinced so many judges, how it took so long for there to be a media backlash, etc.
Apparently there's a UK legal presumption that computer evidence is reliable. (As of Jan 2025, this is under review.) It's also possible some Fujitsu expert witnesses may eventually be charged with perjury. The analogy of a DNA testing lab giving false positives seems a fair one.
While nobody "wanted" to hurt the postmasters, plenty had strong incentives to prefer an individual conviction to an admission that their expensive IT system was fundamentally unreliable, or (eventually) that they had been responsible for sending innocent people to prison (or worse).
Finally, the UK justice system is indeed very slow at fixing mistakes when they do happen. Other recent examples: Andrew Malkinson (after 17 years), Tom Hayes (10 years), Peter Sullivan (38 years), maybe Lucy Letby.
I proudly understand what a type species is.
If you figured it out solely from Wikipedia's definition, I say you've earned that pride.
I picked it up from your quoted excerpt, but I was already familiar with the general concept from outside biology. That said, I might be misunderstanding it entirely and just didn't realize the excerpt wasn't what I expected it to be :D An example like "the grey wolf (canis lupus) will always be the species to which relative genetic distance determined membership in the family."