Hacker News new | past | comments | ask | show | jobs | submit login
My PhD Genealogy (stanford.edu)
129 points by hkc on Oct 19, 2022 | hide | past | favorite | 59 comments



I spent some time in 02016 digging through different sorts of academic lineages. It turns out, for example, that you can also trace Leibniz back to Copernicus: https://dercuano.github.io/notes/academic-lineage.html

Thrun's page seems to have an error about Leibniz: "Gottfried Wilhelm Leibniz 1966, 1967, 1976"

It would be nice to be able to trace figures like al-Tusi back to Plato and Imhotep, to know if there really was an unbroken line of personal mentorship the way there is in the Buddhist lineages, or if at some point the oral line was severed. Perhaps during the Roman rampages through Greece, the line of transmission of philosophy only survived in Alexandria, or less plausibly, somewhere in India, only to resurface in Arabia while Europe was sunken into its Dark Ages. Or perhaps it had to be recovered from the few manuscripts the Christians hadn't yet recycled into hymnals, like the Archimedes Palimpsest.

We know that somewhere between Eudoxus and Galileo the idea of freely postulated axiom systems was lost, and it was not really fully rediscovered until the 19th century.


I feel that the possibility of tracing academic lineage back to antiquity in the West is very dim, for the same reason that tracing descent from antiquity[0] in Europe has proven impossible - too few records survived. Even in the Catholic Church, the longest unbroken chain (i.e., for which records survive) of apostolic succession (i.e., which bishop consecrated each bishop) goes only back to the 1400s with Guillaume d'Estouteville, even though France in the 1400s was long after the Dark Ages and many records survive from the High Middle Ages onward.

[0] https://en.wikipedia.org/wiki/Descent_from_antiquity [1] https://en.wikipedia.org/wiki/Guillaume_d%27Estouteville


A thing I was thinking was that, though almost surely the line was broken in the West, it might not have been broken altogether. We know, for example, that the Pauliṣa Siddhānta in India contains Hellenistic astronomy; and we know that as far back as Classical Athens, Greek philosophers reported visiting India and studying with "gymnosophists", an ascetic tradition (called Digambara in Sanskrit) which either survives there today or has been coincidentally reinvented in the same subcontinent at least a thousand years ago. Surely it is conceivable that, during the period when Rome was laying waste to the Hellenistic world, some philosophers might have fled to India and trained their successors there?

But it does seem very unlikely that we could trace it, given how little written material survives from that period in India.


> Even in the Catholic Church, the longest unbroken chain (i.e., for which records survive) of apostolic succession (i.e., which bishop consecrated each bishop) goes only back to the 1400s

Our (Episcopal) parish used to have a small framed "genealogy" that purported to trace our diocesan bishop's consecration lineage back to St. Peter. I was always a bit skeptical.

Tangent: Some Roman Catholics would flatly deny the validity of any Anglican ordinations post-Henry VIII ....

https://en.wikipedia.org/wiki/Apostolicae_curae


I’ve always wondered what things we’ve lost from library fires, hall burnings, church burnings etc. perhaps this is one such example


It boggles the mind

The likelihood of a text surviving tends towards zero given enough time

Before antiquity we often only have quoted fragments to look at

The early episodes of the History of Philosophy Without Any Gaps is a pretty good introduction to how little we know about pre-socratic thinkers especially

Surviving corpora can often be only hundreds of characters long


I mostly try to be amazed at what has survived (and the unlikely ways it did).

One thing to note is that there seems to be a compounding difficulty of preserving records. Eg we don’t have much science or mathematics from anything like late antiquity but somehow the Hagia Sophia was built in 537. If you look at the evidence that people were still capable of the kind of analysis required to build that structure it seems plausible that much mathematics or science was still happening (in the eastern Roman Empire) but not being recorded. Perhaps if you are copying books you’re more likely to be copying the important ancient foundational texts rather than more advanced narrower more recent things. There’s like 7-800 years between Apollonius and the construction of that church. I do wonder what mathematics was done in that time but not recorded. But it also may be that the culture around mathematics was somehow limited with no (evidence of anything like) algebra or calculus or coordinate systems which are pretty important to the developments in the last few hundred years of mathematics.

Early western Christianity was pretty bad for preserving ancient texts (a silly and likely exaggerated way to phrase the attitude is something like ‘if it’s compatible with the bible it is unnecessary and if it disagrees with the bible it’s heresy and should be destroyed’ though it also seems the New Testament is somewhat Aristotelian) but the eastern empire kept many of them going (in their original Greek), and the Muslims ended up with Arabic translations and Western Europe then got copies or contemporary Arabic works through what is now Spain and translated them into Latin (this could be somewhat tricky for mathematics but worse for anything more philosophical which would have likely been changed first to be compatible with Islam and second to be compatible with Roman Catholicism).[1] Eventually scholars in the west also got access to the Greek texts from the Byzantine empire and the attitudes towards censorship had changed (I guess it was less necessary for things that were not in the vernacular) and it was perhaps lucky that this happened before the fall of Constantinople.

[1] apart from destruction there are also things that are lost due to there being a lot of texts and not that many people looking at them. Eg I think there’s a big library associated with the bishop of toledo (maybe a cathedral or seminary library; maybe in a bishops palace) which is full of Arabic texts from before the rechristianization of Spain. There may be texts there which are unknown to modern scholars.


Agreed.


Can you provide a reference for the claim that the concept of a freely postulated axiom system was implicit in Eudoxus? I'm very curious!

I assume the 19th century rediscovery you refer to was Boole, Hamilton et al and their work in logic and the beginnings of abstract algebra.


I think it's in Lucio Rosso's The Forgotten Revolution but I don't have the page number yet.


Haven't found it yet...


no where on the site does it tell us that a lineage actually is. Does it mean that the mathematicians knew eachoter? that they adviced? cited?


usually it means that the “parent” was a phd supervisor for the “child” researcher who studied for a PhD under that supervisor


> "02016"

Thank you for preparing for the Y10K problem.


How does it help to create a Y100k problem by using fixed-width numerals with five digits, instead of using variable-width numerals like we normally do for years?


exactly, it seems like a faulty solution to a problem that doesn't exist -- at present, we would say 900 and 1, not 0900 and 0001.


I believe the point is to try to get people to engage in long-term thinking.


It's a Long Now thing.


Kurzgesagt suggests we soon have a merry 12023

https://www.youtube.com/watch?v=CWu29PRCUvQ


Academia is not that big, is quite incestuous, and such trees or descent are not rare. I think there was a website that listed people's academic family

here is one for neuroscience: https://neurotree.org/neurotree/


This is the site that does it for math: https://www.genealogy.math.ndsu.nodak.edu/


what do you mean by "incestuous"? I'm having a hard time picturing the tree structure. Like if we keep things simple and assume every person on the tree "asexually spawns" multiple students, where does the "incest" come in?

Maybe one student having more than one advisor? If that's the case, usually it's just a thesis committee or reviewer or something, and not really multiple _main_ advisors


The academia is far more diverse than people think. More diverse between countries, between universities, between disciplines, and over time.

Having a junior and a senior advisor is fairly common. Sometimes the work is done in multiple institutions, with a separate advisor in each. In some systems, most people who supervise students are not formally qualified to do so, creating a need for a separate formal supervisor. Sometimes there are two equal advisors, and sometimes the advisor changes for various reasons. Sometimes the student is an independent scholar and the advisors are only loosely involved in the work. If you only have written records, it can be impossible to tell which of these was the case for a particular student.


> what do you mean by "incestuous"?

To give you an example, about half of my group at University X did a Bachelor's in statistics at University X, then a Mater's in statistics at University X, then a PhD in statistics at University X, and some are even doing a PostDoc (in statistics at University X)!


but that's not what the original comment mentioning "incestuous" was talking about. The original was talking about academic trees, for which "incestuous" would indicate a tree with loops (i.e. multiple paths between two points)


I started browsing https://www.genealogy.math.ndsu.nodak.edu/ after reading this post. Some of the genealogies are wildly impressive. For example:

https://www.genealogy.math.ndsu.nodak.edu/id.php?id=38586

Bernoulli -> Euler -> Lagrange -> Poisson and Fourier


The Euler -> Lagrange link is sort of dodgy, though. Lagrange was an autodidact. What he got from Euler was Euler's position at court, not a doctorate, though they did collaborate on some research before that.


You can really see the results of the ebbs and flows of generations in this genealogy. The massive post-war three part economic, technological, and population booms just jumps out of this data. Ph.D's granted in this genalogy, per decade. 1870's: 1 1880's: 2 1890's: 0 1900's: 3 1910's: 0 (they were fighting a war, no time for dissertations) 1920's: 2 1930's: 2 1940's: 1 (They were fighting a war, no time for dissertations) 1950's: 0 (Baby bust from the great depression) 1960's: 4 (The intra-war and postwar kids, with funding and jobs) 1970's: 2 1980's: 0 (Too soon to be have their own students for a 1995 Ph.D)


This makes me want to get a PhD to join the family.


This looked like fun so I wrote a tiny script to generate these graphs for mathematicians using data from https://www.mathgenealogy.org/. Here is what it looks like for Gauss: https://raw.githubusercontent.com/parsiad/math-genealogy-gra.... If you want to make your own, I stuck the script here: https://github.com/parsiad/math-genealogy-graph.


There is a site for mathemitics where you can search with names: https://www.mathgenealogy.org/ pretty comprehensive according to my anecdotal evidence.


I really like the simplicity of this.

I suppose this was done by hand. Having such an overview while doing the research would be really beneficial for discovering novel ideas and connections. I haven't come across such a tool as of yet.


Much of this information is available from the Mathematics Geneology Project. https://www.genealogy.math.ndsu.nodak.edu/id.php?id=100193


There is academictree.org, which provides "academic genealogy" for quite a few disciplines–is this the type of tool you're looking for?


Thank you! I had never looked up my academic genealogy before, which includes Rosalind Franklin, Linus Pauling, and Michael Polanyi.


I would love to have an ideas and people genealogy where you can select a thinker like Rousseau and have a graph of the main ideas in his works and their predecessors.

Think about his amour de soi. Did it existed previously anywhere else? Who talked about something similar earlier?

I'd die for something like that.


I wanted something like this as well, and I have a prototype of something much simpler (using node2vec to generate embeddings using data from Wikidata and DBpedia (and Twitter)). It doesn't really do what you want, but you might find it interesting.

Rousseau: https://pov.is/e/93f9822c-1ed8-4bc9-aec9-064e7bb6807c Amour de soi: https://pov.is/e/82e9f674-ebbf-4c36-b225-ec1653ce3367

You can go backwards and forwards in time using by-year view (though missing data in Wikidata makes this a bit difficult): https://pov.is/e/93f9822c-1ed8-4bc9-aec9-064e7bb6807c?i=Q5&o...


This reminded of a book that I read a while ago. It listed the important ideas that lead to modern computer science in a chronological manner[1]. Is there any work done to trace the genealogy of computer science like [2]?

[1] https://mitpress.mit.edu/9780262045308/ideas-that-created-th... [2] https://www.genealogy.math.ndsu.nodak.edu/


Depending on your definition of `Computer Science', [2] is already the link you're looking for. Church, Turing, von Neumann, McCarthy, Shannon and many other founding fathers of Computer Science are there. In fact, from your link, I believe the only people mentioned who are not in the Mathematics Genealogy Project are Lovelace and Boole, and that's because the PhD system didn't really exist in England at the time, so a specifically Computer Science based database wouldn't help with that.


If you like this, you might be interested in AstroGen, the Astronomy Genealogy Project that I helped design: https://astrogen.aas.org


thanks for the tip - just added myself!


Jean-Baptiste Fourier

       |
Joseph Lagrange

       |
Leonhard Euler 1726

       |                        
Johann Bernoulli 1694

       |                             
Gottfried Leibniz 1666

I was never aware of this connection. Is there some reasons or story behind how all of these geniuses clustered together?


It's hard to succeed in academia if you aren't in a lineage.

See Castelvecchi, D. "Majority of mathematicians hail from just 24 scientific ‘families’". Nature 537, 20–21 (2016). https://doi.org/10.1038/nature.2016.20491

https://www.nature.com/articles/nature.2016.20491


Same goes with classical piano, and Kung Fu.


Here are some simple visualizations of this data: https://gist.github.com/breck7/b88155a58cbf83d32283b9ea50bf8...


https://academictree.org/flytree/ does something similar for genetics or the base site for dozens of other disciplines.


I will actually give this one a tremendous amount of credit for picking up both my real advisor, and my "shadow" advisor (a very, very close member who was on my committee, but at a different institution).


Leibniz - Bernoulli - Bernoulli - Euler - LaGrange was a pretty good run!


He just needs to make a movie with Kevin Bacon and publish with Erdos to round things out. https://en.wikipedia.org/wiki/Erd%C5%91s%E2%80%93Bacon_numbe...


Ha! Erdős–Bacon numbers originated in a 2002 Telegraph article by Simon Singh, referenced by Wikipedia as [1]:

https://web.archive.org/web/20121112081753/http://www.telegr...

I came up with this game after realizing I had an Erdős–Bacon number. He credits me, but spun the article to make this sound like a "thing" other people cared about.

I was written out of that Wikipedia page long ago. Very little of what survives on that page stands up to close scrutiny.

Combining these numbers is an obscure amusement, but people take the separate numbers seriously. For Erdős numbers, should one count posthumous papers? My "2" via Persi Diaconis goes to "3" if one doesn't.

The original intent of the Bacon number game was to count actors in fictional speaking roles. My "2" here is from a speaking role in "A Beautiful Mind". The "Oracle of Bacon" replaced this intent with whatever their database could easily report. Appearing as oneself in a documentary on Erdős had the obvious hilarious effect.

One understand these links better by studying IMBD credits. Daniel Kleitman's "2" depends on a "Thanks" credit from "Good Will Hunting", and few of the other low Bacon numbers on the Erdős–Bacon Wikipedia page can be confirmed at all.



Erdős has been reading Proofs from the Book for a couple decades, and is unable to publish.

Perhaps some of these people are available:

https://en.wikipedia.org/wiki/List_of_people_by_Erdős_number


Pretty cool. I had never thought to do this we my PhD genealogy before, it took about 10 clicks to get to Poisson (the first name I recognized).


This is very cool, but I think it is a stretch to say that Rheticus was a student of Copernicus.


Do people use these genealogies to pick which schools to go to and people to learn from?


The field of Neuroscience tends to track lineages with Neurotree.org


we are cousins!


neat, I can trace mine to Kant


Sadly a predominantly white male genealogy.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: