Argumentation design in decision support: effects on legitimacy, usefulness and meaningfulness
Decision support is the science and associated set of practices that consist in developing and implementing tools and methods to make use of scientific knowledge to help a decision-maker or a group of decision-makers improve their decisions or the processes through which they make decisions. The exercise of decision support is typically accompanied by, and in some cases even mainly consists in, developing argumentations intended for decision-makers, stakeholders potentially impacted by the latter’s decisions, or the larger public. Such argumentations can take various forms: explanations of how tools and methods are designed and work, justifications of their use in different kinds of contexts, discourses illustrating their properties and impacts, etc. These argumentations can, and in some cases are intended to, play a crucial role in determining how legitimate, useful or meaningful decision support, and in turn supported decisions, are. In the wake of a first symposium organized at LAMSADE (UMR 7243), Paris, in December 2023 and devoted to explore the concepts of legitimacy, usefulness and meaningfulness of decision support, this symposium will explore the scientific and normative stakes of argumentation design in this decision support context. In this 2-day long symposium, mornings will be devoted to plenary talks exposing concepts, and afternoons to concrete discussions of case-studies.
Program
Day 1: Thursday 7 November 2024 – Argumentation in decision support: what it is, and how we use it
- Morning
- Keynote speaker (9:30-10:30) ► Olivier Cailloux (LAMSADE), Wassila Ouerdane (Centrale-Supélec), Gabriella Pigozzi (LAMSADE) : Arguments, explainability, legitimacy: what, why, how?
- Coffee break (10:30-11)
- Contributions (11-13)
- Sarra Tajouri (LAMSADE) : Explanations vs. justifications of algorithmic decisions
- Sara Bianchi (University of Brescia) : How to measure the sustainability of urban transformations in a meaningful and useful way
- Louise Dupuis (LAMSADE) : Abstract argumentation applied to controversies over contested projects: the case of the Saint-Agnan “Mega Sawmill”
- Evan Skinner (Australian National University) : TBA
- Lunch (13-14:30)
- Afternoon
- Case study: presentation (14 :30-15 :30) ► Beatrice MECCA (Politecnico di Torino) : Argumentation schemes for urban decision support
- Coffee break (15:30-16)
- Case study: discussion (16-17:30)
Day 2: Friday 8 November 2024 – Transformative decision support: argumentation and its impacts
- Morning
- Keynote (9-10) ► Bruno Ambroise (Université Paris 1) : La théorie économique peut-elle être performative ? Quelques remarques critiques
- Coffee break (10-10:30)
- Contributions (10:30-12:30)
- Blandine Schaffner (Université Strasbourg) : Implementation of a decision support process for the creation of an ecological corridor in Ivory Coast
- Juliette Rouchier (LAMSADE) : Can it be a mistake to crackdown environmental opposition?
- Maryam Loftian (HEIG-VD) : A Multi-Criteria Decision Support Tool to Highlight the Role of Ponds/Pondscapes as Nature-Based Solutions
- Chabane Mazri (INERIS) : Strengths and weaknesses of the EU Common Safety Indicators (CSI) system
- Lunch (12:30-14)
- Afternoon
- Case study: presentation (14-15) ► Florent Joerin (HEIG-VD) : Conflit, participation et argumentation autour d’une nouvelle infrastructure routière
- Coffee break (15-15:30)
- Case study: discussion (15:30-17)
Abstracts
Ambroise, Bruno (ISJPS, CNRS)
La théorie économique peut-elle être performative ? Quelques remarques critiques
Le concept de « performatif » a récemment gagné un usage en sociologie économique : depuis une quinzaine d’année, plusieurs publications importantes recourent massivement à la « performativité » pour en faire un usage censé permettre de saisir certains phénomènes sociaux spécifiquement économiques. Ces études, prenant le plus souvent appui sur l’anthropologie des sciences et des techniques développées par M. Callon et B. Latour, ou sur les études de D. MacKenzie, recourent au concept de « performatif » pour comprendre comment certaines théories ou modélisations économiques parviennent à construire le réel qu’elles prétendent décrire, notamment des marchés financiers d’un certain type. Ce concept permettrait ainsi de comprendre comment l’économie ne se borne pas à décrire le monde économique, mais contribue aussi à le construire. Tel est le principe – et l’enjeu – d’une école qu’on a pu qualifier de « sociologie performativiste ».
L’ambition – limitée – de cet exposé sera d’explorer l’usage qui est fait du concept de performativité, afin d’examiner si, d’un point de vue épistémologique, il est adéquat aux objectifs poursuivis et correspond à l’ambition explicative qu’il sert. Car s’il est probable que les travaux mentionnés mettent au jour des processus intéressants et importants, et que l’économie comme science (economics) puisse affecter ou modifier, d’une manière ou d’une autre, son objet, à savoir l’économie comme réalité (economy), il n’est simplement pas sûr qu’il s’agisse là d’une quelconque forme de « performativité », et il n’est donc pas sûr que l’usage du concept ne serve pas à travestir d’autres formes d’efficacité, dont la compréhension gagnerait à ce qu’elles soient qualifiées autrement. Réciproquement, il n’est pas sûr que l’usage – même modifié – du concept permette de penser (ou de rendre compte de) ce qu’il est censé identifier selon les tenants de la sociologie performativiste.
Bianchi, Sara (University of Brescia)
How to measure the sustainability of urban transformations in a meaningful and useful way
From the international to the local contexts, many challenges have arisen in establishing the right framework for measuring sustainable development at various scales. This has directly affected the lack of sustainability standards, within the local communities’ regulations, able to guide the Urban Planning practice in a holistic way. The ex-ante sustainability assessment of urban regeneration plans can be an important tool that, through the adoption of specific indicator systems, can: guide the transformation of urbanized areas through sustainability criteria, and support the Public Administration in making informed decisions. However, from a Decision-Aiding perspective, many analysts have questioned the validity of information provided by indicator systems, as well as the pertinence of conclusions driven by their use. During the adoption of the urban sustainability evaluation tools, the formal properties of the indicators involved have not been largely discussed through the lenses of the Measurement Theory and Value Theory. Therefore, this contribution aims at investigating the meaningfulness and usefulness of inferences that can be drawn from adopting various measurement and value scales when assessing the sustainability of urban regeneration plans. In particular, some critical considerations will be made on the use of specific environmental indicators found in the literature of Urban Studies, i.e., the Impervious Surface Coverage, and the Ecological Value and Ecological Balance. The discussion will help in understanding that: (i) the use of quantitative (numerical) information does not guarantee rigor or objectivity, and (ii) the meaningfulness and usefulness properties are not strictly dependent. A decision support methodology is required in order to define the appropriate and rigorous use of measurement systems for evaluation purposes in support of the Public Administration’s decision-making.
Cailloux, Olivier (LAMSADE), Ouerdane, Wassila (Centrale-Supélec), Pigozzi, Gabriella (LAMSADE)
Arguments, explainability, legitimacy: what, why, how?
We will present and contrast a few conceptions of argumentation, justification, explainability and legitimacy that are considered in the literature. We will then illustrate with a few recently proposed concepts and cases. We will discuss an illustrative case in the paradigm of formal argumentation theory and recall the notion of floating conclusions, which may be considered legitimate despite the lack of common justification. We will contrast the former concepts with the one of deliberation thanks to the notion of deliberated judgments: those that a subject forms when faced with counter-arguments. We will end by discussing how the dialectical nature of argumentation matches the explainability feature within a decision support activity.
Dupuis, Louise (LAMSADE)
Abstract argumentation applied to controversies over contested projects: the case of the Saint-Agnan “Mega Sawmill”
In France, the acronym GPII (“Grands Projets Inutiles et Imposés”) refers to major infrastructure projects that are strongly opposed by local populations. Controversies linked to these infrastructures involve a wide range of stakeholders, including activists, the media, elected representatives and industrial players. This work explores abstract argumentation as a tool for analyzing these debates. Focusing on a specific case - the construction of an industrial sawmill in Saint-Agnan, in the department of Tarn, France - this research examines how argumentation graphs can model interactions between opposing camps and reveal the main argumentation strategies used. This work demonstrates how this formalism can be used not only to study but also to influence the dynamics of such debates, offering a method that could be applied to larger, more complex controversies in the future.
Joerin, Florent (HEIG-VD)
Conflit, participation et argumentation autour d’une nouvelle infrastructure routière
En 2018, afin de mettre en oeuvre un grand projet stratégique d’aménagement du territoire, le Canton de Genève projette la construction d’une nouvelle infrastructure routière. Rapidement, celui-ci fait l’objet d’une forte mobilisation d’opposition impliquant des voisins, des agriculteurs, des associations environnementales, ainsi que quelques entreprises privées.
Dans ce contexte, la HEIG-VD a réalisé un diagnostic social, puis animé une démarche participative afin d’accompagner les services publics du Canton et l’ensemble des acteurs concernés vers la recherche d’une solution acceptable. Ce processus a eu pour conséquence de grandement modifier le projet et il semble aujourd’hui mieux répondre à la diversité des attentes. Cette intervention propose ainsi un regard sur l’évolution des arguments ou enjeux portés par les différents acteurs impliqués.
Loftian, Maryam (HEIG-VD)
A Multi-Criteria Decision Support Tool to Highlight the Role of Ponds/Pondscapes as Nature-Based Solutions
Multi-criteria decision analysis (MCDA) can be as simple as comparing two housing options based on sale price and distance to the city center, or as complex as identifying the best locations for installing solar panels, using a wide range of qualitative and quantitative criteria with varying weights. MCDA comes with its own set of challenges, including but not limited to the selection of an appropriate analysis method, difficulty in criteria selection, handling conflicting criteria, and interpreting results. However, one often underestimated challenge in MCDA arises when it is applied to scientific analysis, particularly when the tool is designed for multi-stakeholder use.
In this presentation, we explore the development of a decision support tool within the context of the PONDERFUL project and discuss the challenges faced throughout the process. The PONDERFUL project investigates the role of ponds and pondscapes as nature-based solutions (NBS), considering indicators of nature's contributions to people (NCP), such as biodiversity and GHG emissions. A decision support tool was developed to utilize the models produced by PONDERFUL researchers to evaluate and compare various NBS implementations, such as pond creation or water quality management, based on their impact on NCP indicators. The tool was initially intended to create an interface between decision-makers and the scientific outcomes of the Ponderful project, while also aiming to convince them of the benefits of developing a network of ponds in their region.
We will present the tool we have developed and discuss the specific challenges encountered. In particular, we examine how to simplify MCDA without losing essential information for the analysis, while ensuring the tool remains easy and relevant for decision-makers to use. Additionally, we explore how to effectively communicate the final outcomes in a way that aligns with the project's scientific objectives (support the development of ponds). Indeed, the quality of the scientific results is not always adapted to the decision-making context under consideration (global vs local) and some of these results could produce a confusing message for decision-makers. Understanding these challenges when using MCDA in a scientific context can help pave the way for future implementations and foster a common understanding between researchers and decision-makers.
Mazri, Chabane (INERIS)
Strengths and weaknesses of the EU Common Safety Indicators (CSI) system
High risk systems (Industry, transportation, energy…) are subject to scrutiny in the public area given their potential to trigger catastrophic consequences. National and EU authorities therefore heavily rely on performance indicators and indexes to track trends and detect, as early as possible, inflexions in those systems performances that may lead to adverse events. These information are considered as central as they motivate the annual classification of countries performances and may lead to (re)define national investment policies for the railway infrastructures.
The presentation will focus on a study conducted to analyze the strengths and weaknesses of the EU Common Safety Indicators (CSI) system and discusses more particularly the adequacy of these metrics with the span of uses associated to them.
Mecca, Beatrice (Politecnico di Torino), Lami, Isabella (Politecnico di Torino), Bottero, Marta (Politecnico di Torino), Pigozzi, Gabriella (LAMSADE), Tsoukias, Alexis (LAMSADE)
Argumentation schemes for urban decision support