The TL;DR is that I've been influenced by three main lines of intellectual descent.
Of course it's easy to rationalize post hoc, but all this feels to me like it makes good sense: I'm descended from a combination of mathematics (well, computation, but the two go hand in hand; my undergrad class entering in Fall 1983 was the first one that could declare CS as a major, rather than studying it as a part of applied math), and language, and empirically driven natural sciences. Not just one, but all three. That's pretty cool.
Student | Advisor(s) | Dissertation Year | University | Dissertation Title | Brief notes | |
---|---|---|---|---|---|---|
Philip Resnik | Aravind Krishna Joshi | 1993 | University of Pennsylvania | Selection and Information : A Class-Based Approach to Lexical Relationships | Although Aravind was my official advisor, I was lucky enough to have what I think of as an amazing "advisory cloud" at Penn that also included Mitch Marcus, Mark Steedman, Bonnie Webber, Mark Liberman, and Lila Gleitman. (Plus of course the many incredible student colleagues.) Lila was in many ways my true mentor in terms of her influence on how I think about research, and she's a main reason that I emerged from Penn not only as an applications-oriented NLP person but also a cognitive scientist, so I include her below. I was also influenced as an undergraduate by Bill Woods (important figure in early NLP; dissertation titled "Semantics for a Question Answering System", 1968), who was also later my boss at Sun Microsystems Labs. Woods was advised by Susumu Kuno (dissertation titled "Automatic Syntactic Analysis of English", 1964) and with Kuno he was a reader for my undergraduate honors thesis (which was kindly supervised by linguist John Whitman well outside his comfort zone). Via Woods and Kuno my intellectual lineage goes back to Nathaniel Bowditch, "an early American mathematician remembered for his work on ocean navigation ... often credited as the founder of modern maritime navigation". | |
Lila Gleitman | Zellig Sabbetai Harris | 1967 | University of Pennsylvania | Key figure in cognitive science and especially the study of child language acquisition; co-founded UPenn's Institute for Research in Cognitive Science. Although I list Harris here as her advisor, her obit in the New York Times notes that as she followed Chomsky's intellectual break with Harris it led to "a split so bitter that he refused to oversee her dissertation". (I'm not sure who officially oversaw her thesis; if you happen to, please let me know!) | ||
Aravind Krishna Joshi | Seymour Sherman | 1960 | University of Pennsylvania | Some Applications of Information Theory and a Generalization of Normalized Entropy | Key figure in computational linguistics; co-founded UPenn's Institute for Research in Cognitive Science. I (along with a lot of others, I think) believed that Zelig Harris was Joshi's advisor. However David Chiang dug into this, actually inter-library-loaning Joshi's dissertation, and found that the advisor of record was actually Seymour Sherman. I followed up with Mitch Marcus to get the fuller story. In a nutshell, Mitch reports that Aravind's early work was entirely about information theory/communication theory and, except for one paper in 1958, all of his papers until 1964 were in that area. But then, influenced by Harris, Aravind abandoned information theory per se and pursued the computational work that his interaction with Harris led to. Mitch notes that the birth of cognitive science at Penn was the interaction among the grad students working for Harris, and not Harris himself --- Aravind, Lila, and Carol Chomsky (who did her undergrad at Penn and her PhD at Harvard). | |
Seymour Sherman | John Adam Fitz Randolph | 1940 | Cornell | A Comparison of Linear Measures in the Plane | Academic grandchild of David Hilbert (!), which takes the ancestry back to Carl Friedrich Gauss (!!), and from there back to Leibniz (!!!). | |
Zellig Sabbetai Harris | James Alan Montgomery | 1934 | University of Pennsylvania | A Grammar of the Phoenician Language | Key figure in structural linguistics and more generally the "mathematical and empirical foundations of the science of language" (Wikipedia). Strongly influenced Noam Chomsky, including convincing him to study linguistics as an undergrad. | |
James Alan Montgomery | Hermann Vollrat Hilprecht | 1904 | University of Pennsylvania | The Samaritans; The Earliest Jewish Sect; Their History, Theology and Literature | Clergyman, Oriental scholar, and biblical scholar | |
Hermann Vollrat Hilprecht | Friedrich Delitzsch and Heinrich Leberecht Fleischer | 1883 | Universität Leipzig | Freibrief Nebukadnezar's I, Königs von Babylonien c. 1130 v. Chr. zum ersten Mal veröffentlicht, umschrieben und übersetzt | German-American Assyriologist and archaeologist, also with a healthy dose of theology. | |
Heinrich Leberecht Fleischer | Unknown | 1824 | Universität Leipzig | De Glossis Habichtianis In Quatuor Priores Tomos MI Noctium: Dissertatio Critica | Specialist in theology and "oriental languages" (Arabic, Turkish, Persian); founded Arabic Studies in Germany. Interesting quote from Wikipedia that strikes me as anticipating a more scientific approach to linguistics: "The question for us is not: What is the purest, the most beautiful and correct Arabic, but what is Arabic in general?" | |
Friedrich Delitzsch | Franz Julius Delitzsch | 1873 | Universität Leipzig | Studien über indogermanisch-semitische Wurzelverwandtschaft (Studies on Indo-European-Semitic root relationship) | Deeply anti-semitic German nationalist who specialized in the study of ancient Middle Eastern languages | |
Franz Julius Delitzsch | Unknown | 1836 | Universität Leipzig | From Wikipedia: "Delitzsch is considered to be one of the most important Old Testament exegetes of the 19th century. Comprehensively taught, a connoisseur of rabbinical literature, an exegete who combined biblical-theological interpretation with philological meticulousness." |
Ok, now, here is where it gets wild. David Chiang pointed out that academictree.org is way more complete. Here are some interesting highlights of my academic ancestry from info there, working backward in time. The tree widget appears below and you can use the "P+" button to go backward in time (and then when it gets too big scroll bars for horizontal and vertical scrolling will show up). Or, to see the whole thing, here's a PDF constructed thanks to David's graciously pointing me to this code.