Past Fellowship Recipients and Projects

Lital Belinko Sabah Hebrew University of Jerusalem
2017-2018

The present project involves an analysis of what is traditionally known as the present tense in folk narrative texts written in Judeo- Spanish, also known as Ladino. The corpus from which I plan to draw my data consists of a collection of 85 stories from the Balkans that were transcribed and published between 1914-1935 by different Romanists, philologists and folklorists.

Berenice Darwich​ City College, CUNY, and Lehman College, CUNY
​2018-2019

The purpose of this investigation is to account for the distribution of él and ella (‘he, she’) in Spanish. This project builds on, and refines, previous research (Darwich 2016) on the presence and absence of these forms in co-referential clauses, that is, adjacent clauses where same reference is maintained in verbs that are in the closest proximity. Using naturalistic data from the PRESEEA-Mexico City corpus, I analyze the communicative strategies of use of every instance of él and ella(and not only in co-referential environments) in order to update their hypothesized meanings in earlier CS works: Entity in Focus, Person, High-Deixis, Gender, Number (García 1975, 1983, 1996 and 2009).

Kelli Hesseltine
​2016 - 2019, 2021-2023 City College of New York

Kelli’s previous linguistic research, conducted with her mentor Joseph Davis, involved an examination of the order of words traditionally classified as nouns and adjectives.  Over the course of three years, Hesseltine and Davis developed and gathered data for the Assertion of Characterization hypothesis, exploring two different meanings - WEAKER (signaled by the order AB, with a preposed adjective), and STRONGER (BA, postposed) - with which language users communicate an assertion of an entity’s characterizing traits.  In 2020, their work, “The communicative function of adjective-noun order in English,” was published in the academic linguistic journal WORD and there they identified several issues for future consideration, one of which is Hesseltine’s research focus for this year’s fellowship. 

Eduardo Ho-Fernández​ City College, CUNY

2017-2022

Building on previous Society-sponsored research, this project investigates a complex set of word order facts in Spanish involving event words and words inferred to be participants in those events. The hypothesis is that position before the event (PEP) signals HIGHER, and position after the event (PEP) signals LOWER, Participant Attentionworthiness.
The current project will analyze one-participant events, developing a hypothesis where EP and  PE) are considered signals of HIGHER or LOWER Event Attentionworthiness.
An important feature of this project is the treatment of EPP and PPE. Preliminary analysis suggests that they are manifestations of the same signaling mechanism hypothesized for EP and PE (i.e., EPP and PPE). As such, these sequences do not represent the signals of a different grammatical system. A challenge stemming from this proposal that will be addressed in the current project is that a residual participant (P), conveying no grammatical status, would be left over in both EP(P) and (P)PE, its distribution requiring an explanation.

Aaron Liebman 2020-2021

This fellowship project builds on research that Aaron Liebman conducted many years ago as part of an undergraduate course with Dr. Alan Huffman, in which he collected examples of the English 'bare infinitive' (infinitives without 'to', from Harper Lee's novel 'To Kill a Mockingbird'). Mr. Liebman wrote a paper, presented his findings, and discussed his examples with William Diver himself and with other Columbia School linguists at the Friday morning seminar. That early contact began an interest and association (including attending summer workshops at CUNY and Rutgers and later proofreading papers for a published volume) which has continued to this day. Mr. Liebman hopes to contribute to the validation of the Probability or Likelihood System originally proposed by William Diver and developed and used for decades as a working hypothesis for English by Aaron's friend and mentor Alan Huffman.

Verónica Mailhes Universidad de La Plata, Argentina
2017-2018 Graduate Student Scholarship Recipient

This proposal aims to investigate, following the principles of Columbia School Linguistics, the variable use of so-called future tenses in Spanish, conventionally known in the grammars as the morphological future (FM) and the periphrastic future (FP). In a project of wide scope, we start by limiting our data to political discourse in River Plate Spanish. The project offers first a critique of the meanings that have been traditionally attributed to these forms: higher vs. lower degree of facticity of the event. In this proposal, it is postulated that objective facticity is not the relevant semantic substance. Rather, we explore the possibility that the relevant substance is control by the speaker over the event. When politicians want to express control over the event, they opt for the PF form, opting for the MF form where less control is involved. Take note of the semantic substance involved. Under this hypothesis, the term 'control' refers to control by the speaker, not control by a participant or controller in the event named by the verb, as in existing analyses of Spanish clitics, Latin case endings, and English word order.​

Andrew McCormick  City College, CUNY 2017-2019

This project concerns the word how in English. Traditional analyses have posited at least three categories for how: manner adverb (‘how to make lasagna’), intensifier (‘how far/many’) and complementizer (this role is applied to situations where how is believed to be interpreted like ‘that,’ as in, ‘He said how nobody likes the new policy’). Adopting Columbia School’s monosemic bias (Ruhl 1989, Reid 2004), this project will seek to demonstrate that there is only one how, and its context-invariant meaning pertains to the relevance of elaborating information.  
This analysis will be shown to account for several problems with traditional approaches that posit these discrete categories. One is that some tokens of how are ambiguous between manner adverb and complementizer (Willis 2007, van Gelderen 2013). Another is that some tokens that best fit the ‘complementizer’ category appear to have a mild sense of degree (van Gelderen 2013 11-12; cf. Greenbaum 1996: 146). Finally, some tokens are even triply ambiguous, allowing the manner, intensifier or complementizer interpretations. In brief, the extent to which these categories overlap raises questions about their reliability, and suggests that there may be only one how.

Employing Columbia School’s strict distinction between meaning and message (Reid 2018, Stern 2019), this project will argue that the traditional categories summarized above reflect messages – contextually dependent interpretations. These messages all fit how’s more abstract, stable meaning: that elaborating/characterizing information is relevant to the utterance and its surrounding discourse. 

Data will be drawn from the Corpus of Contemporary American English (COCA; Davies 2008- ) and at least one full-length novel (O’Brien 1991).

Max Miller The City College of New York, CUNY

This project studied the two forms that indicate PAST Time in English—PAST (e.g. chose) and PAST, BEFORE (had chosen)—and their relative meanings. In traditional grammar, there was an assumption that the PAST, BEFORE is used to narratively link two events in time, placing one before or after another. However, the PAST has been seen to perform the same function (Vonnegut 1973). The Columbia School has shown that a difference in signals reliably indicates a distinction in meaning. What distinction does a language-user mean to indicate when using either the PAST or the PAST, BEFORE to link two events? In previous CS treatments (Huffman 1989, Reid 1978), the PAST, BEFORE has been proposed as a form reserved for events of less importance, or focus. However, strictly speaking, this distinction lies outside the domain of Time, that is, when events occur. Therefore, we will consider the possibility of the distinction between the two forms being within or without the grammatical system of Time. Data was drawn from the Kurt Vonnegut novel Breakfast of Champions, which had already been proven to be a rich source of the grammatical phenomenon of interest.

Roxana Risco National University of La Plata, Argentina
The problem concerns the study of what has been called preposition A in Spanish. Traditional grammars have described it as part of a closed-class word group or as a functional morpheme, positing a series of semantically related meanings or senses for it. However, the attempts at explicating the semantic contribution of A that would account for its distribution still seems to be an ongoing task. Why, for instance, do speakers use A in: Juan fue a París (John went to Paris), Juan vio a María (John saw María), Juan vive a café con leche (John lives on caffe latte), or Juan conoce a todo el mundo (John knows everyone) vs. Juan conoce todo el mundo (John knows the whole world)? The goal is to explain what may lead Spanish speakers to produce A in certain contexts in the particular way they do. Also, of main interest to my research is the possibility of a monosemic analysis of A. Based on observable data and theoretically in line with the analytical sign-oriented tradition of the Columbia School of Linguistics, this study relies on the explanation for the use of A in terms of a meaning hypothesis: a single invariant semantic value functioning as a communicative tool that simply contributes to the communication of many different types of messages or message partials rather than encoded messages.

Nadav Sabar Hebrew University of Jerusalem, 2017-2018

Post Doctoral Fellowship Recipient
The problem investigated concerned the distributions of the forms all, each and every. These forms are all analyzed by logicians, formal semanticists and cognitive linguists as English language representations of the logical universal quantifier. Yet despite this general agreement regarding quantificational force, subtle differences among these forms have long been acknowledged, and attempts at explicating the precise semantic contribution of each in a way that would account for each form’s peculiar distribution is still an ongoing task. Why, for example, do speakers use all with both Entity-s and Entity-0 (e.g., all times and all time, respectively) while each and every are used virtually only with Entity-0 (e.g., every time); or why does the sequence all together occur significantly more frequently than the sequence all individually though both sequences are found to occur many times? Of interest also is the question regarding the identity of the signal in certain high frequency combinations. For example, is all in all right or all of a sudden the same linguistic unit all as found in, say, we all worked on it? Tentative hypotheses for the meanings of all, each and every will be tested through both qualitative analyses of attested examples as well as large scale quantitative predictions carried out in the Corpus of Contemporary American English. Tying in to my previous meaning analysis of some and any, the goal of this project will be to provide a comprehensive sign-based account of the distributional problems posed by all, each, every, any and some, forms which have all been analyzed as logical quantifiers of one sort or another.

Daniël Pieter van Soeren University of Groningen 2018-2019

The communicative role of lexical stress in Columbia School Phonology
This research studies the distribution of phonemes within Spanish CVCVCV words and English and Dutch CVCVC words from a Columbia School perspective (Diver 1979; Tobin 1997), and focuses on lexical stress as a relevant factor in the distribution of apical and visible phonemes. We argue that word recognition is key in the explanation of the results; in the pronunciation of a word, every phoneme eliminates a number of word candidates, making the identification of the word progressively easier. At the beginning of the word, where the word is least predictable, there is a higher burden on distinctiveness, which results in a favouring of other kinds of phonemes than at the end of the word. 

Studies on word recognition in Spanish, English and Dutch (e.g. Cutler 2005; Soto-Faraco, Sebastián-Gallés & Cutler 2001; Van Donselaar, Koster & Cutler 2005) suggest that in addition to segmental information, the position of stress eliminates word candidates too (e.g. a word-initial stressed syllable /ˈru:/ could result in rumour, but eliminates routine as a candidate). In Spanish, English and Dutch discourse, the majority of words begin with a stressed syllable. Consequently, an unstressed initial syllable eliminates more word candidates, which affects the distribution of the aforementioned phonemes.

Lauren Whitty

My fellowship is focused on the investigation of must, may and could. This research builds on my previous research where I used a Columbia School lens to understand the distribution of can (and could). My hypothesis for can is part of a monosemic system of meaning which accounts for all occurrences of can, and identifies two forms of could (one being what is traditionally described as the past of can, and the other having its own place in the modal, or likelihood, system).

In the present investigation, I continue to explore may, might and the “other” could as part of this system. Based on my previous project, my very initial hypothesis is that these lexical items are used when a speaker assesses the likelihood of, but does not completely affirm, an utterance, with a scale whereby must is its highest member, may is in the middle,

and could is its lowest member.
Data will be drawn from the naturally occurring data of recent written texts (for example, Owens’ (2018) Where the Crawdads Sing). Texts as such provide rich, expanded context for interpretation, help elucidate the relationship between lexical items, as well as demonstrate the need for analyses to go beyond the sentence level to understand the communicative function of must, may and could. Furthermore, findings will be tested in the written portion of the Corpus of Contemporary American English (Davies 2008-).
Through continuing my research on this system, I am building on Diver’s (1964) previous work. By approaching these terms synchronically, and with a Columbia School lens, I am hopeful to one day have a completed investigation of the whole system and account for the forms can, could, could, must, may, might, as well as the other forms that may be included (e.g. should, might).

Davies, M. (2008-). The Corpus of Contemporary American English: 450 million words, 1990-2015. Retrieved from http://corpus.byu.edu/coca/
Diver, W. (1964). The Modal System of the English Verb. Word, 20(3), 322–352. Owens, D. (2018). Where the Crawdads Sing. New York: G.P. Putnam’s Sons.

Past Seed Grant Recipients

2022-2023 - Lorena Martín, Eberhard Karls Universität Tübingen and Olga Steriopolo, Leibniz-Zentrum Allgemeine Sprachwissenschaft (ZAS)