Hadas Kotek » Teaching

Click on the blurb link to see a description of the course in-line, or on the syllabus link to download a PDF copy. In some cases, I additionally include a link to course materials.

Demystifying Large Language Models [Fall 2024]

Special seminar (undergrad/grad), MIT. Blurb, Syllabus

This course explores the abilities and limitations of language models, focusing on state of the art tools such as GPT-4, Gemini, and LLaMA. Large Language Models (LLMs) possess impressive language abilities, but they also occasionally fail in unpredictable ways. Our goal in this class will be to map the abilities and limitations of these models, focusing on complex reasoning and language abilities. We will attempt to discover systematicity in the models’ failures and to understand how they relate on the one hand to how the prompt is formulated and what we believe the training data and model architecture to be, and on the other hand how humans perform on the same tasks and how children acquire this knowledge. We will additionally entertain the various costs associated with the deployment and use of LLMs, be they due to privacy breaches, environmental costs, security risks, copyrights abuses, the environment, or the entrenchment and amplification of biases and stereotypes at scale. Along the way, we will investigate the development of language technologies and their capacities over time, as well as the state of the art linguistic theories that explain the phenomena of interest. We’ll ask ourselves whether it is reasonable to conclude that the LLMs use a similar sort of approach as humans do to complex language reasoning, and what this means for how we should understand what LLMs actually do (and how humans can and should interact with them).

Exploring Language Technologies (aka Demystifying LLMs) [Fall 2023]

Special seminar (undergrad/grad), MIT. Blurb, Syllabus

This course explores the abilities and limitations of language models, focusing on state of the art tools such as ChatGPT. LLMs possess impressive language abilities, but they also occasionally fail in unpredictable ways. Our goal in this class will be to map the abilities and limitations of these models, focusing on complex reasoning and language abilities. We will attempt to discover systematicity in the models’ failures and to understand how they relate on the one hand to how the prompt is formulated and what we believe the training data and model architecture to be, and on the other hand how humans perform on the same tasks and how children acquire this knowledge. Along the way, we will learn about the development of language technologies and their capacities over time, as well as the state of the art linguistic theories that explain the phenomena of interest. We’ll ask ourselves whether LLMs appear to resemble humans in their approach to language and reasoning, and what this means for how we should understand what LLMs actually do (and how humans can and should interact with them).

Illusions of Language [Spring 2017, Spring 2019]

Freshman seminar, Yale University. Blurb, Syllabus

In order to understand language, speakers must successfully navigate complex linguistic expressions on a rapid time-scale. Failures of the linguistic parser provide a window into how language is encoded in the brain and how it is deciphered in real time. This course gives special attention to ’grammatical illusions’: expressions which the parser mistakenly accepts as grammatical, despite making little sense upon closer reflection, and on the other hand grammatical sentences which the parser has difficulty processing. Emphasis is also put on illusions and misconceptions about language, such as the belief that women speak more than men, that ‘vocal fry’ can harm your voice, and that double negation as in I don’t have no time for this is illogical.

Syntax 2 [Spring 2019]

Advanced syntax for undergraduate and graduate students, Yale University. Blurb, Syllabus

A key topic in the theory of syntax is that operations apply in limited domains. These involve restrictions on argument selection, agreement, syntactic binding, and movement. This class aims to explore a range of syntactic phenomena, accounts proposed for them, and their theoretical implications, especially in relation to recent minimalist conceptions regarding the design of the language faculty. We concentrate in particular on locality constraints on agreement and movement, presented through the lens of the contemporary probe-goal approach to Agree. We will use this backdrop to further develop the theory of grammar established in Syntax I.

The phenomena to be discussed will relate in particular to (a) how the syntactic (computational, structure building) component of grammar interacts with, and is limited by, the systems it interfaces with (specifically, PF and LF), and (b) parameters of grammar assumed to capture cross-linguistic variation.

Theoretical topics addressed will include (at least some aspects of) the following:

the operation Merge: phrase structure and movement;
wh-movement, superiority, and islands;
units of the syntactic derivation: phases/cycles and their size;
the operation Agree: phi-features, feature-checking, probes and goals, (defective) intervention.

Research methods [Fall 2018]

Research methods for undergraduate seniors, Yale University. Blurb, Syllabus

This course is for seniors who major in Linguistics, and has two main goals. The first one is to train you to do research in linguistics, and facilitate the process of writing your senior essay. This means that your main tasks will be to take the necessary steps that will lead you to produce your own piece of research, which overall include

identifying a research area,
identifying a research topic within this area, and an advisor,
acquiring the background knowledge you need,
formulating a research question,
formulating a hypothesis,
testing your hypothesis,
examining and evaluating the results.

The second goal is to help you develop professional skills that are useful for your future, such as writing abstracts, research statements and proposals more generally.

Semantics 1 [Fall 2016, Fall 2018]

Introduction to semantics for undergraduate and graduate students, Yale University. Blurb, Syllabus

Introduction to truth-conditional compositional semantics. Set theory, first- and higher-order logic, and the lambda calculus as they relate to the study of natural language meaning. Some attention to analyzing the meanings of tense/aspect markers, adverbs, and modals.

Special Topics: The Syntax of Ellipsis [Winter 2015, Spring 2018]

Undergraduate seminar, McGill University (2015) and NYU (2018). Blurb, Syllabus, Materials

Under certain conditions, linguistic material can be omitted from an utterance but we nevertheless understand the utterance as if it has not been [ ]. The challenge of associating meanings with “silence” has made this [ ] a central topic in linguistic research. In this course, we will explore a variety of phenomena that involve omission of *[ ], mainly involving ellipsis in the verbal domain. We will ask ourselves not only what material can be elided, but also why [ ]; and we will discuss at length whether it’s possible to “rescue” some ungrammatical structures using ellipsis, and if so, which ones [ ]. We will study both seminal and current papers in the syntactic literature, and also ?[ ] psycholinguistic aspects of the processing of ellipsis and its perception. Students in the class will be required to do weekly readings, actively participate in *[ ], give one in-class presentation, and write a short squib.

Language [=Introduction to Linguistics, Spring 2018]

Introduction to Linguistics, New York University. Blurb, Syllabus

This class will provide some answers to basic questions about the nature of human language. Throughout the course, we will be examining a number of ways in which human language is a complex but law-governed mental system. We will study some core aspects of this system in detail, and then use what we have learned to address a variety of other questions – including how language is acquired, how dialects arise, how languages change over time, and others. The main component of evaluation will be linguistic problem solving through assignments and exams.

Wh-constructions cross-linguistically [Fall 2017]

Graduate syntax seminar, New York University. Blurb, Syllabus

This seminar will investigate the syntax/semantics of diverse wh-constructions cross-linguistically. The first half or so of the seminar will be devoted to questions, based in part on my forthcoming LI monograph, and providing familiarity with modern analyses of interrogative syntax/semantics. The second part of the seminar will seek to situate the machinery developed in the first part of the seminar in the broader context of other A-bar and scope-taking phenomena, as well as a study of a variety of other wh-constructions, including (at least some of) relative clauses and free relatives, wh-indefinites, wh-quantification, free choice, and wh-NPIs.

Grammatical Analysis [=Syntax 1, Fall 2017]

Undergraduate introduction to syntax, New York University. Blurb, Syllabus

“A logical theory may be tested by its capacity for dealing with puzzles, and it is a wholesome plan, in thinking about logic, to stock the mind with as many puzzles as possible, since these serve much the same purpose as is served by experiments in physical science.” Russell (1905, 484–485)

This course is an undergraduate-level introduction to modern syntactic theory. Empirically, the course will cover a range of grammatical phenomena in English and other languages, to “stock the mind with as many puzzles as possible,” and to motivate and evaluate hypotheses under discussion. Students will complete the class with both the technical expertise and theoretical foundation to approach a range of work in contemporary syntactic literature. Course requirements include problem sets roughly every other week, two exams, and in-class participation.

Semantics 2 [Spring 2017]

Advanced course in semantics for undergraduate and graduate students, Yale University. Blurb, Syllabus

This course is the second course in sequence with LING 263/663, which is a prerequisite for this course. The goal is to expand students’ understanding of the empirical domain and techniques in formal semantics and pragmatics as well as to develop skills in doing independent semantic analysis. We will focus on analyzing a range of empirical phenomena and use these to strengthen our feel for asking questions and solving puzzles in Semantics. This course has a heavy hands-on component. Early in the semester, each student will come up with a topic that they want to work on in more detail. The topics covered and readings done in the second part of the semester will depend on student choices for research topics. I will designate readings around these topics and they will be read by all of us so that each student can “tinker” around fruitfully with the appropriate semantic tools to come up with an analysis.

Questions and focus [Fall 2016]

Undergraduate/graduate seminar on the syntax and semantics of questions and focus, Yale University. See Fall 2014 for materials. Syllabus

Introduction to the Study of Language [Winter 2016]

Introduction to linguistics for non-majors, McGill University. Blurb, Syllabus

Introduction to Linguistics [Winter 2016]

Introduction to linguistics for majors, McGill University. Blurb, Syllabus

This course is an introduction to linguistic theory and analysis. Topics covered include phonetics, phonology, morphology, syntax, semantics and pragmatics. The goal is to familiarize students with current tools of linguis- tic analysis. Students will be required to display an understanding of these tools by applying them to data from different languages. The main component of evaluation will be linguistic problem solving through assignments, quizzes and exams.

Questions, focus, and friends [Fall 2014]

Graduate seminar, McGill University. Blurb, Syllabus, Materials

In this seminar we will explore the syntax and semantics of questions and focus constructions. From a theoretical point of view, we will discuss in detail two technologies used for scope taking—(covert) movement and focus alternative computation—which are commonly employed in the analysis of both questions and focus constructions. From a more typological perspective, we will explore the shared overt morphosyntactic strategies some languages use in the expression of both kinds of constructions.

Phenomena to be discussed include in-situ and ex-situ wh-questions and Association with Focus constructions, pied-piping, movement asymmetries and islands, intervention effects, and alternative questions. Time permitting, we may discuss other phenomena for which both (covert) movement and alternative computation have been (or could be) employed, such as disjunction, NPIs, universal quantification, and head-internal relatives.

Requirements for registered students will include infrequent homework assignments and two language journals, which report on the investigation of wh-questions and focus constructions in a particular language, based on elicitation with a native speaker. We will assume some familiarity with properties of A’ (wh) movement and (extensional) compositional semantics as in Heim & Kratzer (1998), but important parts of the theory will be reviewed in class. Pass/fail registration is welcome.

This seminar led to a Winter semester reading group on wh-indefinites: Materials.

Turkshop [Spring 2013]

Workshop at the Experimental Syntax/Semantics lab at MIT department of Linguistics Blurb Materials from MIT workshop (2013); Materials from NYU tutorial (2017)

Workshop on the use of Amazon’s Mechanical Turk, introducing turktools: a set of tools designed to assist researchers to create a wide range of linguistic tasks, including linguistic grammaticality surveys, sentence completion tasks, and picture-matching tasks. The tools further help streamline the design of such linguistic experiment and assist in the extraction and analysis of the resulting data.

###