Domestic water uses from a culturalist approach

David Chaquea Romero

23 October 2025

By Tima Miroshnichenko from Pexels

Overview

This contribution focuses on the culturalist approach of the Social Practice Theory, highlighting that practice should be the ontological unit of analysis when investigating residential water consumption. Water-use practices greatly influence current daily water consumption patterns, hence also offering the greatest potential for change. Ultimately, a better understanding of current water-use practices will lead to a smoother transition to a less-consumptive lifestyle. This is demonstrated with the case study of the PUDA2022 project (Domestic Water-Use Practices).

PUDA 2022 Project being applied in Cali, Colombia

Case study

This project is currently carried out in the city of Cali, Colombia (circa 3 million inhabitants), working with a randomised representative sample of 382 households within the city. In these households, the PUDA2022 (Domestic Water-Use Practices Survey) has been conducted with any household member older than 16 years old. This questionnaire was designed based on 90 informal conversations, with various individuals, from 33 households (September 2020).

The survey consists of two sections:

social-demographic and economic characterisation of both, the general household and each participating household member;
the characterisation of six domestic water-use practices:

Washing Clothes
Body Washing
Toilet Use
Handwashing
Brushing Teeth
Dishwashing

This characterisation was made based on Shove, Pantzar, and Watson’s (2012) research, who categorised water-use practices into three sub-elements: materials, competencies, and meanings.

Furthermore, an important characteristic of this questionnaire is that it recognises that each water-use practice element could be labelled at varying levels of recurrence. Therefore, the response options have been categorised as Always, Most of the Time, Rarely or Never, as later seen in the figure - Possible different multilevel networks.

Characterisation of domestic water-use practices

Following the work of Medd and Shove, 2006 and Pullinger et al., 2013, a k-means clustering technique has been applied to identify widely shared “versions” of a practice. This was first done using data from a Pilot Test, to ensure its workings. Identified clusters were plotted using Rings of Recurrence radial graphs, developed and programmed by the project’s researchers. This tool helps visualise the differences between various practices, as shown in Rings of recurrence radial graphs.

Shove et al. (2012) highlighted the need for a dynamic index of social transformations in practices, to see how practices have changed over time. Using this outlook, for this study it is important to know how stable, fixed, and/or resistant to change the water-use practices we surveyed are. To calculate the degree of stability of a cluster of practice a Social Stability of Practice Index is proposed. Fuzzy Logic was selected as the software to integrate the multiple data inputs, which are categorised into Endogenous Properties (the isolated cluster) and Exogenous Relationships (the cluster and its household environment).

Considering that practice is defined as a network of interconnected elements, it is important to highlight that the Fuzzy Logic calculation of SSPI is focused mainly on elements (nodes) in relation to their recurrence. With this in mind, as well as the nodes, links are an essential part of the networks. To visualise and calculate the properties of the assemblies, we explored network diagramming and analysis. In this stage, we realised that we had multiple levels of recurrence, this means that each node could have three different positions: Always, Most of the Time and Rarely, and it could be linked with the other elements of those three categories. In this sense, it is possible to identify 6 networks (as shown under).

The six identified networks

Rings of recurrence radial graphs

Watch David explain how he uses this method in his research

To identify these multilevel links, an adjacency matrix was developed cross-referencing data using functions in an Excel worksheet like this =COUNTIFS(A1:An;”ALWAYS”;B1:Bn;”MOSTOFTHETIME”). This function counts how many times the link between two variables appears fulfilling the two conditions. Using the software Gephi, following Higginson et al., (2015), it was possible to model the different multilevel networks that provide structures to the practice.

Hierarchical Fuzzy System

Each network has its own properties, some having stronger or weaker links than others, affecting the stability of the practice. We think that Fuzzy Logic calculations and Network Analysis can be useful to understand the always-changing structure of the practices, in our case, domestic water-use practice. This kind of knowledge could be important to recognise the complex structures we are dealing with and to develop strategies to ignite the rearrangement of the structures of practice. As well as, to foster new ways to encourage sustainable water-use and its patterns of consumption.

Possible different multilevel networks

This PhD research is being developed in the Environmental Sciences Doctorate program in Universidad del Valle, Cali, Colombia; and is part of the Water Security and Sustainable Development Hub funded by the Global Challenges Research Fund.