Philosophy of Language and Computation, Autumn 2022
ETH Zürich: Course catalog
This graduate class, taught like a seminar, is designed to help you understand the philosophical underpinnings of modern work in natural language processing (NLP), most of which is centered around deep learning applied to natural language data. The course is a year-long journey, but the second half (Spring 2023) does not depend on the first (Fall 2022) and thus either half may be taken independently. In each semester, we divide the class time into three modules. Each module focuses on a philosophical topic. In the first semester we will discuss logic and language, structuralism, and recursive structure. And, in the second semester we will focus on language games, information theory, and pragmatics. The modules will be four weeks long. During the first two weeks of a module, we will read and discuss original texts and supplementary criticism. During the second two weeks, we will read recent NLP papers and discuss how the authors of those works are building on philosophical insights into our conception of language—perhaps implicitly or unwittingly.
The course will require a weekly reading of around 20 pages (with relatively high variance at times) and a weekly free-form responses to the reading, which is to be completed online. The mandatory readings for every week are bolded in the syllabus below. The weekly tasks are short and not graded, but, in order to pass the class, at least 70% of the tasks must be completed. See this document for the guidelines on the free-form responses. The final grade will be based, in equal parts, on one class presentation and two term papers (each around 5-10 pages) which are to be turned in at the latest one week before the start of the exam period, that is, on Sunday, January 15th, at midnight. Students must write term papers on the modules other than the one for which they have made a presentation. They will be expected to explore the relation of the topics discussed in class to work not presented in the class, focusing on the connection between the philosophy of language and NLP. For example, discussing how three recent NLP papers implicitly assumed a structuralist perspective on language would be a good topic.
Marks for the course will be determined by the following formula:
- 1 / 3 Class presentation
- 1 / 3 Term paper 1
- 1 / 3 Term paper 2
More detailed instructions about the paper and presentation requirements can be found here:
Tue 16-18, HG E 1.1. The lectures will be given in person. This recurring Zoom meeting will be used throughout the semester for people who want to tune in remotely. The password can be found on the course Moodle page. The recordings of the lectures can be found on the course Polybox and the password can be found on the course Moodle page as well.
Discussion Sections: Tue 18-19 HG E 1.1.
Course Live Chat: To provide an easier way to communicate with the teaching team and your colleagues, we have set up a course chat server. We use RocketChat and the server is hosted on ETH servers. We encourage you to sign up and participate in the discussions there.
Important: There are three important points you should keep in mind about the course live chat:
- RocketChat will be the main communications hub for the course. You are responsible for receiving all messages broadcast in the RocketChat.
- Your username should be
firstname.lastname. This is required as we will only allow enrolled students to participate in the chat and we will remove users which we cannot validate.
- We highly recommend you download the desktop app here.
07.08 Class website is online!
28.09 Presentation assignments have been published.
29.09 The course Polybox has been published. You can find the password on the course Moodle page.
Syllabus and Schedule
You can ask questions on the course chat server. Please post questions there, so others can see them and join the discussion. If you have questions which are not of general interest, please don’t hesitate to contact us directly, i.e., email the lecturers with the TAs cc-ed.