Seeing the point of politics
Exploring the use of 'argument visualisation' techniques as aids to understanding the content of political debates in the Scottish Parliament
MSc: Software Engineering, 2004
The practice of using software to 'visualise' the structure of arguments is used profitably in academic and business circles. Insofar as political issues may be classified as 'Wicked' problems, it is reasonable to assume that ones understanding of a political debate might be served by the presence of such an argument visualisation (also referred to as an 'argument map'). The purpose of the project is to assess the validity of such an assumption.
Research background: Technical
Prose is an imperfect medium for conveying information. Grammatical rules impose words on most sentences, yet these words add little in the way of content. Where a piece of text extends over a number of pages, the proportion of uninformative material increases. Thus, when faced with an analytical task, such as tracing a line of argument, the analyst must learn to strip away the verbiage that encumbers the path they are attempting to follow.
Argument Visualisation is a technique that assists in this process. It goes back at least as far as the early 20th century, when diagrams were used to depict the arguments contained within legal texts as an aid to law students. Legal arguments were found to be easier to appreciate if they were set out as an arrangement of boxes and arrows, rather than as prose. Developing a series of icons to represent parts of an argument, such as 'rebuttal', 'claim' and 'warrant', further decreased the amount of text required to illustrate the case. This practice spread to the commercial world, being adapted to meet the particular needs of business.
By the early Sixties, computers had developed to the point where it was possible to couple their ability to organise and manipulate data with the benefits of diagrammatic representation of material, resulting in greater efficiency in exploring problems and more fruitful ways of developing solutions. This was enhanced in the following decade by research into the nature of the problems faced, in particular, by the commercial sector. Horst Rittel and Max Webber distinguished between problems that were 'Tame' and those that were 'Wicked'. The former are not necessarily trivial, but occupy domains with an established practice for finding solutions, which is often so systematic that it is possible for the process to be automated. 'Wicked' problems can be characterised as follows: they are difficult to formulate; they lack clear solution strategies; they generate further problems as the attempted solution progresses; and at no point can one say with confidence that a solution has been reached. Rittel and Webber concluded that such problems were best tackled using argumentation techniques, and developed 'Information Based Issue Systems' (IBIS) as a consequence.
Supplementing IBIS with hypertext functionality enables the user to access a wealth of additional material without obscuring the main picture of the argument. Instead of having a picture cluttered with both claims and their associated warrants, the latter can be provided virtually through hyperlinks to documents or further argument maps, thereby maintaining clarity and integrity.
To date, many different software tools exist for creating argument maps. The maps in this project were created using 'Compendium'.
Research background: Political
Over the last few British elections, there has been a disturbing decrease in voter turnout. The results of surveys support the belief that the electorate is becoming increasingly disillusioned with the processes of government. However, there is evidence to show that in those constituencies where parties have taken the trouble to provide information about their policies, there has been a corresponding increase in voting. Hence there is an incentive to investigate new ways of communicating the activities of Parliament in a way that re-engages the public's interest.
One obvious facet of political life is the debate, where different policies are contested during the creation of legislation. Taking the Scottish Parliament as an example, the public can follow a debate live, by visiting the speaking chamber or following the web broadcast, or they can read the Official Report that is published by the Parliament, both online and as a paper copy. Yet, each means of access presents obstacles to a lucid understanding of what is taking place. By attending a live debate, the spectator faces the difficulty of maintaining a clear idea of past comments with which to compare any later observations. Looking at the Official Report overcomes this problem but there remains the impediment of searching back and forth through the pages in order to pursue a particular line of thought.
Two debates were selected for visualisation in order to evaluate the utility of CSAV techniques applied to politics. One was a short debate on TETRA masts. This was chosen as an instance of a topic that had important health concerns for the population as well as implications for the efficiency of the emergency services, but which contained off-putting technical jargon. One was a long debate, stretching over two sessions, of the high-profile issue concerning Antisocial Behaviour, which raised many separate though related issues.
Argument Map creation
Maps were created on the basis that the following findings were broadly accurate: that people were more likely to vote when they had a clear understanding of the issues at stake; that people felt disengaged from their MSP outwith elections, and were cynical about their representative's performance once elected. Accordingly, the maps conveyed information on the issues that were debated, as well as providing facts about the MSPs who took part.
Each map commenced with a node containing the motion of the debate, along with the MSP who moved it and the MSP who responded to the motion. The remainder of the responses were collated into groups of issues, and arranged on the map in a logical fashion.
The maps had the following features:
- Icons to represent the main aspects of debate:
- Comments in support of a position.
- Comments that contest a position.
- References alluded to in a comment
- MSP Information:
- MSP's name, party and constituency - made available through cursor rollover.
- Summary of MSP's comment - held as text in an icon.
- Debate information:
- Statements grouped together according to topic.
- Lists of comments by individual and by party.
- Lists of suggestions or questions that were not responded to in the debate.
- Embedded maps to divide the debate into manageable portions.
TETRA argument map.
Navigation is managed using the vertical and horizontal scroll bars.
Placing the cursor on the blue asterisks in the top right hand corner of ions reveals information about the politician(s) associated with the text.
Left-clicking the mouse key when the cursor changes from an arrow reveals further information.
The project lacked the resources required to determine the extent to which these maps could convert someone from an uninterested state into a politically active member of society. Instead, it aimed at providing evidence to support a case for continuing this line of enquiry. To this end, a structured interview was arranged with the Director of the Scottish Civic Forum, and an informal interview was conducted with two individuals representing a voter's perspective on the matter. In addition, the Scottish Parliament Information Centre (SPICe) provided a report upon the usefulness of the maps from a politician's perspective.
The participants were provided with argument maps of the debates accompanied by access to the relevant Official Report, and were encouraged to comment upon the content and format of the visualisations, as well as make suggestions for additional use. A summary of these responses was prepared and submitted to the participants for approval. The following is a very brief account of their views.
Director Scottish Civic Forum:
For researchers, although the Official Report is an essential document, the maps provide a powerful overview of the issues canvassed, especially with the division of comments into topics. The impersonal presentation of debate is beneficial insofar as it makes for clarity, but doing so sacrifices details concerning inter-party conflict. Because the maps provide an indication of the complexity of a politician's task, the public is more likely to take a sympathetic view of their MSP's position, thereby improving the relationship between the citizen and their representative. Broadly speaking, the maps were judged to have considerable potential.
The Official Report only had one clear advantage over the map, and that was in its preserving the effect produced by a politician in the development of their argument. The participants also noted the equivocal nature of presenting the issues impersonally. Otherwise the maps were regarded as far more stimulating than the report, a useful resource for locating member's opinions on issues, and were viewed as a valuable technique in the pursuit of political openness.
Whilst acknowledging argument visualisation to be an innovative and interesting approach, it was felt that argument maps carried too little factual detail to be useful to MSPs.
Argument Visualisation appears to suit the citizen above all, and to a lesser though still significant extent the researcher. However, the professional politician would not be expected to find them advantageous. A number of observations upon presentation were made, all of which could easily be incorporated in the software. These techniques could be extended beyond the isolated debate to provide a valuable record of a policy's evolution; maps at each stage of the legislative process would increase the level of transparency with which citizens could access the politician's deliberations.
Although the findings are based upon a small sample, they are sufficiently encouraging to warrant taking the research one stage further by introducing the medium to larger numbers of the public.
Links on this website
Each publication is linked to its reference
- Exploiting Argument Mapping Techniques to Support Policy-Making
- Argument Visualisation to support democratic decision-making