Spencer Caplan

I'm a third year Ph.D student (entered 2015) in the Linguistics Department at the University of Pennsylvania where I'm advised by Charles Yang and Mitch Marcus. I am also a member of the Language Development and Language Processing Lab in the Department of Psychology.

Prior to Penn, I completed Undergraduate degrees in Linguistics and Computer Science at Brown University where I worked with Eugene Charniak in the Brown Laboratory for Linguistic Information Processing


My research is at the intersection of linguistics, computation, and cognition.

  • I primarily work on computational models of language acquisition and proccessing, the mental processes that govern how a child acquires whatever language they are exposed to.
  • I apply quantitative methodology (connecting usage to Grammar) to answer questions in theoretical linguistics: What exactly do people "know" when they know a language? How can records of their speech help shed light on that?
  • My other work uses linguistic universals and insights from child language acquisition to improve low-resource NLP.

Code for many projects is available on my Github. If there's something you can't find or can't get to work as expected, feel free to contact me.

I have written for Language Log.

Publications and Presentations


  • [10]   S. Caplan. Word Learning as Category Formation. SCiL 2018, Salt Lake City, UT
  • [9]   S. Caplan & K. Djärv. Embedded-Verb Second is Anti-Licenced by Discourse Familiarity.
  • [a] Meaning in Flux 2017, Yale University, New Haven, CT
    [b] Texas Linguistic Society 17, Austin, TX
    [c] Mid-Atlantic Colloquium of Studies in Meaning 2017 , Georgetown University, Washington D.C.
    [d] SelectionFest 2017 , Berlin, Germany
  • 2017

  • [8]   S. Caplan. Mandarin “Descriptive Complements” are Adjuncts. Texas Linguistic Society 17, Austin, TX (slides|abstract)
  • [7]   S. Caplan & J. Kodner. Vowel Harmony as a Distributional Learning Problem. CogSci 2017, London, UK (poster)
  • [6]   S. Caplan & K. Djärv. Quantitative analysis of Formal and Stylistic variation in Embedded V2. FWAV 4, York, UK (slides|source code)
  • [5]   J. Kodner, S. Caplan, H. Xu, M. Marcus, & C. Yang. Case Studies in the Automatic Characterization of Grammars from Small Wordlists. ComputEL-2, Honolulu, HI (poster|proceedings)
  • [4]   S. Caplan. Morphological Typology Induction. DARPA LORELEI (contact for slides)
  • 2016

  • [3]   {S. Caplan & J. Kodner}. A Computational Model of Vowel Harmony Acquisition. NECPhon10, Amherst, MA (slides)
  • [2]   S. Caplan & J. Kodner. Induction of a Morphological Feature Typology, Automatic Vowel Harmony Characterization, Uyghur A-Raising. DARPA LORELEI (contact for slides)
  • 2015

  • [1]   S. Caplan, N. Fox, M. McClosky, & E. Charniak. Lexical Substitution for Cross-Domain Parser Adaptation. Undergraduate Thesis (contact for paper)


Science happens in conversation, not in monologue. I've had the opportunity to work and collaborate with many smart and interesting people including: Charles Yang, Tony Kroch, Mitch Marcus, Kajsa Djärv, Jordan Kodner, Eugene Charniak, Hongzhi Xu, Neal Fox

Funding, Fellowships, and Scholarships

Editing and Reviewing

Teaching Assistantships


Ballroom Dance

Outside of academia, I am a competitive Ballroom Dancer.

Ballroom dance arose as the standardization of various folk partner dances around the start of the 20th century. Over the subsequent hundred plus years, it has evolved into a formal canon which is very technically-involved and increasingly athletic while retaining the classic (Ballroom) aesthetic.

My dance partner and I are currently ranked as one of the top 50 Amateur Ballroom couples in the US. Competition results and video are available (here) and (here). Our primary coaches are Emanuele Pappacena and Francesca Lazzari. Addtionally, we've had the opportunity to train with many of the top couples internationally.

I also enjoy teaching Ballroom dance, both working with couples individually as well as teaching group classes. I am a Standard Ballroom instructor at the University of Pennsylvania.


My office is 514, Levine Hall.

You can email me at spcaplan at sas dot upenn dot edu.

Miscellaneous Treasures

Avril 50 is one of the finest spots in Philadelphia.

It only took 154 5:00 AM trips and counting to get my photo on the wall at the best restaurant in Providence.

If you find yourself in downtown Austin, you should probably spend all your time and money at Péché.

Everyone should watch more Fellini.