Dieses Kapitel taucht in die komplexe Welt der Designprozesse ein und konzentriert sich auf die Unterscheidung zwischen epistemischen und pragmatischen Handlungen. Es baut auf etablierten Methoden wie Protokollanalyse und Linkographie auf, die verwendet werden, um Designaktivitäten zu studieren und zu analysieren. Die Autoren führen einen erweiterten linkographischen Ansatz ein, um die oft übersehene Unterscheidung zwischen Handlungen, die zu einem neuen Verständnis führen (epistemisch) und solchen, die dem Verständnis folgen (pragmatisch), zu hinterfragen. Aufbauend auf dem theoretischen Rahmen von Kirsh und Maglio stellt die Studie traditionelle lineare Ansichten von Design in Frage und schlägt stattdessen einen zirkular kausalen, dialogischen Prozess vor. Die empirische Laborstudie umfasst Designer, die laut denken, während sie sich mit digitalen Designaufgaben beschäftigen, und liefert reiche qualitative Daten. Die erweiterte Linkographie-Methode ermöglicht eine differenziertere Analyse von Designphasen, die das Wechselspiel zwischen erkenntnistheoretischen und pragmatischen Handlungen offenbart. Dieser Ansatz bietet nicht nur ein tieferes Verständnis von Designprozessen, sondern auch praktische Werkzeuge für Designpraktiker und Forscher, um ihre Methoden zu reflektieren und zu verfeinern. Die Ergebnisse unterstreichen die Komplexität und Interdependenz von Design-Maßnahmen und werfen ein Licht auf die Dynamik von Design-Denken und -Praxis.
KI-Generiert
Diese Zusammenfassung des Fachinhalts wurde mit Hilfe von KI generiert.
Abstract
This paper presents a method for enquiring into a distinction between two kinds of purposeful human action in design: pragmatic and epistemic actions. Although frequently mentioned in the literature, this distinction has not been empirically tested in the design context. To test the distinction, we conducted an empirical lab study with designers thinking-aloud during digital design processes. This study allowed us to collect and subsequently transcribe, parse, and code a set of design process protocols. Based on this protocol set, we tested the distinction between pragmatic and epistemic actions using protocol analysis and a linkographic notation that we extended for this purpose. Through this analysis, we found that pragmatic and epistemic actions are neither all-encompassing nor necessarily mutually exclusive. We thereby arrive at a more nuanced categorisation of purposeful human actions in design as well as an extended linkographic notation, with a potential for more general utility in empirical design research.
1 Background
In design research, protocol analysis and linkography have emerged as well-established methods for studying design processes. Protocol analysis commonly involves subjects verbalising their thought processes aloud, either during or after completing design tasks [1, 2]. Records of these verbalisations are then transcribed into qualitative data for further analysis. Linkography is a widely used notation system for analysing transcribed protocols, primarily focusing on the structured recording and linking of observable designerly actions [3]. In this paper, we present our extension of linkography to capture and test a previously proposed distinction between epistemic and pragmatic actions [4]. While this distinction is frequently mentioned in the literature, it has yet to be tested in light of empirical (digital) design observations [5]. Our methodological approach to investigate this distinction is based on design cybernetics, in which design is construed as a circularly causal conversational process between a designer (the self) and their external environment (the other) [6, pp. 15–18]. In the following sections, we present a theoretical background on epistemic and pragmatic actions, and linkography.
1.1 Epistemic and Pragmatic Actions
The theory we test in the context of design is Kirsh and Maglio’s distinction between pragmatic actions (i.e. actions that follow understanding) and epistemic actions (i.e. actions that lead to new understanding) [4, 7]. With this distinction, the team challenges earlier conceptions of purposeful human action as linearly goal-directed, with planning necessarily preceding execution [4, 5]. Other scholars, such as Schön and Glanville, note that this linear view was prevalent in early design theories, reflecting attempts to prescribe design strategies and actions. More recent theories, however, describe designing as circularly “reflective” or “conversational” [7, 8].
Anzeige
Kirsh and Maglio [4] challenge the above-described earlier notion that “the point of action is always pragmatic” and that ‘understanding’ always precedes ‘acting’, through an empirical study of Tetris play. Tetris players manipulate falling pieces to form and dissolve horizontal rows. Each piece can be moved horizontally and rotated as it falls. The somewhat controlled fall of each piece, and the possible dissolution of horizontal rows caused by it, constitutes a ‘Tetris episode’. During each episode, the preferable target position (or one of multiple, equally preferable positions) for the falling piece, and the shortest path toward it, is unambiguously clear in principle. As the game progresses, however, the speed of falling pieces increases, making it increasingly challenging to identify and attain the ideal position for each piece. This often leads to misplaced pieces and incomplete – and therefore undissolved – rows. The game ends when the screen fills up with incomplete rows, preventing new pieces from entering.
Employing a bespoke implementation of the game, Kirsh and Maglio [4] record actions performed by subjects during Tetris episodes. They observe that players frequently do not move pieces towards their preferable positions directly. Instead, players oftentimes move pieces away from their preferable positions temporarily. This leads Kirsh and Maglio to distinguish two categories of player actions, which they label pragmatic actions and epistemic actions. They describe pragmatic actions as pursuing clearly determined goals directly (see Fig. 1).
Fig. 1.
An exemplar Tetris episode that showcases pragmatic actions.
By contrast, Kirsh and Maglio [4] describe epistemic actions as probing the game environment with a view to improving players’ understanding of possible goals and the actions needed to achieve such goals, as shown in Fig. 2.
Fig. 2.
An exemplar Tetris episode that showcases epistemic actions.
Kirsh and Maglio [4] note that typical Tetris play is characterised by an interplay of epistemic and pragmatic actions and that earlier theories fail to account for the difference between both categories and their interplay. They furthermore claim that their distinction “holds generally throughout all of human activity” [4, p. 548].
Anzeige
Unlike Tetris play, design is characterised by ill-defined, open-ended goals [9] with no pre-defined and pre-determinable units and criteria of analysis. In such cases, Kirsh [10, pp. 252–253] notes that epistemic and pragmatic actions are not directly observable and would require ‘micro-analysis’ and (‘ethnographic’) interpretation. To conduct this type of observation and analysis, we employ linkography – a well-established notation system for investigating the design process at a fine-grained level – to examine epistemic and pragmatic actions in the context of applied design processes.
1.2 Linkography
Introduced in the early 1990s by Goldschmidt, linkography is a notation system to notate and link (protocols of) observable design moves and articulated design ideas in early conceptual phases [1, 11]. Since its inception, the notation has been employed and extended in various research contexts. Given the breadth of related literature, this section focuses on key concepts and extensions of linkography in other studies.
Goldschmidt departs from the earlier assumption “that effective reasoning in a creative endeavour must perforce aim at first mining and then relating to one another the many items of data that are relevant to the task” [12, p. 595]. Linkography thus seeks to parse the design process at a ‘micro-scale’ and represent it quantitatively with a high degree of fidelity [1, 12]. More specifically, linkography segments the design process into consecutive units of analysis – observable designerly articulations and actions – which Goldschmidt refers to as “design moves” [11, p. 291]. Goldschmidt describes a design move as “a step, an act, an operation, which transforms the design situation relative to the state in which it was prior to that move” [13, p. 195], or “an act of reasoning which presents a coherent proposition pertaining to an entity that is being designed” [14, p. 72]. Once identified, design moves are arranged sequentially and linked to form linkographic diagrams. Goldschmidt [1, 15] states that the linking of design moves requires close acquaintance with the task at hand and “common sense”. More specifically, “a link between two moves is established when the two moves pertain to the same, or closely related, subject matter(s), such as a particular component of the designed entity, its properties and functions, a concept or a design strategy, and so on” [15, p. 90]. Design moves and links are visually represented in linkographs, as shown in Fig. 3.
Fig. 3.
An example of a linkograph that features some of its common features and terminology.
In a linkograph, every link can be seen as either a “backlink”, where move B links back to an earlier move A (‘B builds on or responds to A’), or as a “forelink”, where move A connects to a subsequent move B (‘A inspires or prompts B’) [16, p. 385]. Goldschmidt states that aggregated patterns of links offer insights into idea development temporally [1, 17]. For example, a sparse distribution of links might indicate a “sporadic and poorly structured” design phase [17, p. 129], while a denser one might indicate design “fixation and lack of innovation” in another timeframe [2, p. 330].
Linkography is among the few notation systems that originated within, and has been applied beyond, the design research context [3, p. 110]. Some aspects of linkography have been questioned, however, primarily regarding the ambiguity of what constitutes design moves and how links are established among them [3, 18].
Addressing these shortcomings, van der Lugt [18] examines brainstorming processes among groups of designers. He adapts linkography to examine links “between design ideas, rather than design moves” [18, p. 64]. In this approach, design ideas are the units of analysis, which are then notated in spreadsheet-based linkographs [1, p. 165, 18]. Van der Lugt also considers links in the conventional linkographic format as insufficiently differentiated, as they are “solely based on determining the existence or non-existence of connections between elements in the design process” [18, p. 62]. Addressing this lack of nuance, he introduces three kinds of links that offer a more detailed representation of connections among design ideas.
Building on van der Lugt’s work, Hatcher et al. [17] examine design teams during “Design Improv” and brainstorming sessions. They analyse their protocols using van der Lugt’s notions of design ideas and links, which they represent using Goldschmidt’s linkograph notation. In their study, Hatcher et al. consider van der Lugt’s links as insufficiently distinct and, therefore, define more categories of links to analyse their data. They note that their extended linkograph enables “quantitative measures of qualitative transcript data that provided insights into process efficiency, inter-connectivity of ideas, group collaboration, and ideation thinking styles” [17, p. 146].
Linkography has also been extended to encode different design moves [1]. For example, in the context of digital design, Lee and Ostwald [19] employ linkography to “measure cognitive complexity in a parametric design process” among a small group of designers thinking-aloud while using parametric design software. Their design moves (i.e. units of analysis) were encoded based on a coding scheme developed from an existing theory but adapted to examine parametric design workflows. With their adapted linkographic notation, Lee and Ostwald enquire into “cognitive complexity” and note that linkographic patterns are indicative of different design strategies, such as the goal-directed “working-forward strategy” [19, p. 174].
In the following sections, we document our methodological approach and probe the extent to which conventional linkography enables the analysis of less observable, interpreted designerly actions, i.e. epistemic and pragmatic actions.
2 Methods
We conducted an empirical lab study of six (two novice, two intermediate, and two advanced) individual designers concurrently thinking-aloud during digital design processes. We asked the subjects to design a retail shelving and space partitioning system using Grasshopper in Rhino3D, creating a panelled approximation of a double-curved surface on one side and a compartment configuration on the other [5]. Requiring the identification and implementation of suitable geometry rationalisation strategies, this task is open-ended. The design processes of the six subjects were observed remotely via video conferencing software and lasted between 60 and 105 min. We collected qualitative data comprising hand sketches, Rhino3D and Grasshopper files, and audio/video recordings. These were transcribed and divided into two parts: records of observable actions in the external environment (freehand sketching and Rhino3D and Grasshopper interactions) and transcripts of verbalised think-aloud articulations of thought processes [5].
To investigate epistemic and pragmatic actions, we aimed to analytically discern ‘design moves’ (our unit of analysis at a more granular level) and their interconnections in order to discern ‘design episodes’ (our unit of analysis at a coarser level). At the granular level, we encoded design moves based on a coding scheme developed from Schön’s ‘reflective practice’ theory [8]. Our scheme comprises actions starting with a subject’s verbalisation of their setting of an objective, followed by “move experiments” [8] towards fulfilling the stated objective, and followed by a reflection on the extent to which that objective has been fulfilled, leading another cycle with the same pattern [5].
We recorded our encoded design moves in protocol tables along a vertical time axis. We linked these actions using conventional linkographic techniques and probed the extent to which this representation enabled the analysis of epistemic and pragmatic actions.
3 Findings and Extension of Linkography
To fit the scope of this paper, we present here one exemplar partial protocol table and linkograph from our study featuring data representative of what we typically discerned from our design process protocols. As shown on the right-hand side of Fig. 4, the conventional linkographic notation provides quantitative indicators such as forelinks and backlinks that enable us to analyse the interconnections among actions and the development of ideas. At a coarser level, design episodes can be discerned analytically by examining links between actions at more granular levels, such as objective settings and their fulfilments, or vice versa. The conventional linkographic notation, however, lacks the means to gain more nuanced qualitative insights into the nature of these actions and the resulting design episodes.
Fig. 4.
Protocol table and linkograph.
Based on the initial version of this linkograph, we re-examined pertinent verbalisations of design objectives as well as their possible fulfilments in order to categorise each such relationship as either epistemic or pragmatic. In this categorisation process, we exercised our judgment using ‘inference to the best explanation’ [20]. Our ‘evidence’ included qualitative indicators from the transcribed design process protocols and quantitative indicators (i.e. forelinks and backlinks) from the linkographs. If, for example, a subject verbalised an objective to “act in order to see what the action leads to” [7, 8, p. 145], and pursued “exploratory” design moves (‘What if…?’), then we categorised the said objective as epistemic. By contrast, if the subject verbalised an objective to act “in order to produce an intended change” [7, 8, p. 146] and pursued moves accordingly (‘Let’s do x’), then we categorised the objective as pragmatic [5].
Moreover, to categorise fulfilments of these objectives, we examined instances in which the subject reflected on the present state of their observable design work. If, for example, we observed the geometry remaining ‘fixed’ (i.e. to stop evolving) and forming a part of the overall shelving system, then we labelled it as a pragmatic fulfilment (‘That’s good enough, let’s keep it.’). If, however, we observed the geometry to be a snapshot of an ongoing process of enquiry and subject to further re-consideration and change, then we labelled it as an epistemic fulfilment (‘Now I see a good way to proceed.’) [5]. We then revised the linkograph with our descriptive categorisation of each action and extended it further with a more nuanced notational symbol set to represent pragmatic objectives and fulfilments as squares, and epistemic ones as circles (as shown in Fig. 5). We discerned a minimum of 70 and a maximum of 110 design episodes in the six protocol data sets.
Fig. 5.
Revised protocol table and extended linkograph.
Our resulting extended linkograph format shows that epistemic and pragmatic episodes are not necessarily mutually exclusive. We often observe pragmatic design objectives yielding epistemic fulfilments, and vice versa. On some occasions, moreover, episodes turn out to be dead-ends (‘This doesn’t work. Let’s see if there are better ways forward.’) [5]. The design episodes we analyse thus fall into a total of six different possible relationships, starting with either pragmatic or epistemic objectives and ending with either pragmatic or epistemic fulfilments, or dead-ends.
4 Summary and Contributions
In this paper, we detail our methodology to test the epistemic–pragmatic distinction in design. Our lab study involved designers thinking-aloud while engaged in design tasks, from which we collected, transcribed, parsed, and coded qualitative data. The coded actions were then recorded and linked using conventional linkography.
Conventional linkography represents design processes quantitatively and, in our study, enables the identification of design episodes. The linkographic notation system, however, falls short in differentiating less observable designerly actions, particularly when qualitative insights are required. Addressing this, we re-examined our linkographs to categorise objectives and fulfilments as epistemic or pragmatic using our collected qualitative protocols and linkographic forelinks and backlinks. We then revised and extended the linkographs with our qualitative descriptions and with more nuanced notations to represent epistemic and pragmatic actions. Our analytical approach – to re-categorise more granular design actions and coarser design episodes into epistemic or pragmatic – sets our methodology apart from others used in design studies to date.
Our extended linkographs show that epistemic and pragmatic design episodes are not necessarily mutually exclusive. We identify more kinds of episode that Kirsh and Maglio, as well as conventional linkography, do not account for. With this differentiated view of the epistemic-pragmatic distinction, we hope to offer design practitioners insight to reflect on and refine their practice. Similarly, we hope that our extended linkographic notation will be of practical use to empirical design researchers.
Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
