Table for two: Explaining variations in the evaluation of authenticity by restaurant critics

Alessandro Lomi, Anastasia Giachanou, Fabio Crestani, Spyros Angelopoulos

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review


In this paper, we present the preliminary results of a study on the relationship between organizational identity, multiple categorical affiliation and evaluation. More specifically, we are interested in understanding how food critics’ evaluation of restaurants and their judgments of authenticity of individual ingredients vary across national cuisines and across types of restaurants. We want to understand, for example, why “olives” tend to attract positive critical evaluation in the context of Greek restaurants, while they tend to be associated with negative evaluations in the context of Vietnamese cuisine. Also, we want to understand how serving olives would affect critical evaluations of a “Greek-Vietnamese” restaurant.

We explore the conjecture that observed variations in the evaluation of food critics is related to the notion of authenticity, which we reconstruct in terms of affiliation of the restaurants to multiple national cuisines interpreted as institutionalized social categories, and in terms of the relative prevalence of ingredients in each cuisine. The analysis is based on textual data that we have extracted from the corpus of restaurants reviews published in The New Yorker weekly magazine between 2001 and 2013 (called “Table for two”). Starting from a corpus of approximately 300,000 (non-unique) words, we construct three bipartite networks of restaurants-cuisines, ingredients-cuisines, and ingredients-restaurants, which include 512 ingredients (from abalone to zucchini), 42 national cuisines (from American to Vietnamese), and 547 restaurants.

We incorporate a network perspective as well as proximity-based sentiment analysis techniques to assess the evaluation associated with each ingredient mentioned in the reviews. The analysis focuses on: i) the notion of “authenticity” of the restaurants as reconstructed in terms of the number of cuisines to which they are associated and their categorical “contrast”, ii) the extent to which the ingredients associated with a restaurant are “typical” of the cuisines to which the restaurant is affiliated, and iii) how critics’ evaluations about ingredients vary as a function of the kind of restaurant to which the ingredients are associated.

Our work in this paper attempts to pursue three main objectives: i) shed light upon the ways in which food transcends socially constructed dichotomies such as innovative and traditional, local and global, inclusive and distinctive, and, standardized and idiosyncratic; ii) elucidate the ways in which critical discourse about food shapes, manipulates, activates, or neutralizes authenticity as an evaluation schema, and iii) elaborate on the ways in which taste is shaped through organizing and organized discourse.

We discuss the implications of our work for both organization theory and evaluation practice. The results of our study contribute to recent research on organizational identity by showing how relational and categorical identities jointly shape audience evaluations. Finally, we outline future steps on this research project, and elaborate on the research avenues that this work might open for organizational studies.
Original languageEnglish
Title of host publicationProceedings of the 12th Organization Studies Workshop (OSW 2017)
Subtitle of host publicationFood Organizing Matters: Paradoxes, Problems and Potentialities
Publication statusPublished - 2018
EventOrganization Studies Workshop: Food Organizing Matters: Paradoxes, Problems and Potentialities - Chania, Greece
Duration: 18 May 201720 May 2017
Conference number: 12


ConferenceOrganization Studies Workshop
Abbreviated title(OSW 2017)


Dive into the research topics of 'Table for two: Explaining variations in the evaluation of authenticity by restaurant critics'. Together they form a unique fingerprint.

Cite this