Ability to map complex theories

What?

Changes are made to the way that Argument Mapping is normally done to enable one to visually display complex theories clearly and reveal their underlying structure. See Theory Mapping and Epistemology for a brief discussion of the structure of knowledge.


Why?

In outlining complex theories, there are likely to be premises that support more than one argument. For instance, a conspiracy theory is likely to have premises about the extent to which one can trust government that will be repeatedly called upon in justifying arguments for not taking official evidence at face value. The standard practice in Argument Mapping is to keep strictly to a hierarchical approach in which arguments are ‘trees’ where the branches radiate outwards and never connect to each other. So if a premise is used to support more than one argument, it is simply duplicated. The key reason for this is aesthetic, in order to prevent having lines crossing over boxes (e.g. see the article ‘Argument Mapping and Analysis of Competing Hypotheses’ by Tim van Gelder here). It can also make an argument easier to read, since all of the premises are close together. This makes sense for presenting simple arguments, but for Theory Mapping it is better to put each premise in only one place, since: a) any premise has to be justified, and it would be confusing and space-consuming to have to duplicate the justifications too; b) for the measurement of coherence, a premise should only be counted once; and c) if a premise is used in more than one argument, it is good to make it explicit so that the structure of the theory can be more clearly seen, and to assist in weighting premises according to their importance to the theory.

How?

I suggest that the best way to do this while minimizing cross-cutting lines and maximizing clarity, is for a Theory Map to consist of different layers of hierarchical trees (rather than a potentially messy web of connections). The first top layer is for the argument trees justifying reasons and premises that are only used once. The second layer is for any premises that are used twice or more in the first layer. The third layer is for any premises that are used twice or more in the second layer, and so on. The premises and arguments in the lowest layer would therefore be the most fundamental, at the centre of the theory or belief system.

The premises at the top of the second layer would be linked to the arguments in the first layer by lines which would run down the sides of the trees, and so avoid crossing over any boxes. Functionality could also be provided in the software so that if you click on the top of the argument in the first layer, the line down to the premise in the second layer and the premise itself would be highlighted (such as you have with cell formulas in Excel). So it would still be easy to read the argument.

As experience is gained from Theory Mapping about the structure of theories, and software is improved, it may be that even better ways can be developed for presenting maps e.g. where the software automatically structures the map to fit the structure of the theory.
f

Argument Mapping

What?

Argument Mapping is the visual representation of the structure of an logical argument, where premises and conclusions are placed in boxes connected by arrows, such as in the following:

This was done using the bCisive argument mapping software, where the top level conclusion is called a 'Contention', and any arguments or premises supporting it are labelled 'Reasons'. Reasons can either be on their own or part of compound arguments consisting of various co-premises, as shown on the left. There are many different types of box you can have in an argument map, but in the version I use for Theory Mapping, it is kept as simple as possible to just the types of box shown above. In Theory Mapping, every Reason must be supported either by another Reason or by a 'Fact' which is empirical evidence agreed between the debate participants (rather than just what one person considers as 'evidence').

Although I came up with the idea for argument mapping myself a couple of years ago, it turns out to have quite a long history. This is not surprising given how useful it is as a tool to assist human reasoning. Indeed, what is more surprising is why it has not taken off earlier - perhaps because of the lack of software to facilitate it.

Interest in Argument Mapping is now growing rapidly as a way to augment human reasoning for business, personal, academic and educational use, with a number of software applications developed to facilitate its use e.g. bCisive, Rationale, Inspiration, Webspiration, aMap.

A number of websites have also been set up to facilitate debates using collaborative argument maps or text-based versions e.g. Debategraph, Cohere, Deliberatorium, Truthmapping.com, Argunet.

For more information about Argument Mapping, a good place to start is the Austhink website.

Theory Mapping applies the tool of Argument Mapping specifically for the generation, communication and selection of theories (for how this is done, see the other features of Theory Mapping, and the Process of Theory Mapping). This means that it can effectively complement the above online debate websites which are primarily debating about what 'ought' to be done, rather than what 'is'. Being clear about the current situation or best model to describe reality, is a necessary first step before one can make predictions about the future and identify and agree on what action to take.

Why?

Mapping a complex argument has the following potential benefits:

• Clarity of thought: mapping encourages us to express thoughts and their interrelationships much more precisely than we usually do, either when thinking alone or debating with others. By stating claims simply and unambiguously, and by having to draw lines of inference between such claims, we understand the precise nature of the argument much better.
• Uncovering hidden assumptions: having to use logically valid arguments requires that any implicit assumptions are made explicit. Furthermore, requiring that every premise be justified requires continually asking ‘Why is this true?’, leading to uncovering layers of arguments and assumptions that were lying beneath the surface of whatever argument you started with.
• Clear communication: once one is used to reading argument maps and the software is user friendly, they can arguably communicate complex arguments more effectively than prose. The hierarchical structure of argument maps allows the reader to see instantly the ‘gist’ of the argument at the top level of the map. The reader can then drill down the map to see as much detail as they want. Furthermore, all rhetoric and extraneous information is stripped away to only convey the essential logic of a case.
• Effective evaluation: it is easier to identify and comment on weak areas in an argument, whether in terms of the logical structure of the argument, or the justification of the premises. This can be done by inserting Objections into the map, such as the red Objection given in the example map above. Such Objections also need to be supported with reasons and evidence, and can be rebutted.
  

Furthermore, conducting a debate by drafting argument maps of the issues not only benefits from the above advantages over traditional means of debate (verbal, email or internet discussion forums), but also has the following benefit:

• Focus: digression both intentional and unintentional is prevented, since any point has to be addressed to a part of the argument (see the above argument map debating websites for more information about this).

It should be mentioned however, that these benefits are only 'potential'. Given that it is still a relatively new tool, there is currently a lack of empirical research into its effectiveness.
f

Theory Mapping Website

f
The website would include the following:

• Theory Mapping software: to allow you to read, create and modify Maps;
• League tables of theories: for every Research Question, these would give the top scoring theories. Anyone visiting could either accept the ranking and decide to believe the winning theory, or take issue with it and participate in a Debate or Competition to promote their preferred theory.
  
• Discussion Boards: for each Research Question, these would contain a thread for each area of Agreed Facts in which the precise wording of the facts would be discussed by the debating parties and interested onlookers. Incoherencies within each theory would also be discussed, with the Discussion Board recording what was agreed.
  
• Libraries of Maps: libraries of Theory Maps searchable by Research Question, theory, author etc.
  
• Database of Agreed Facts: any facts that were agreed in Debates or Competitions would be stored here, so that they can be referred to and used in the future. This database could be seen as similar to Wikipedia, except that all the content is divided into user-friendly units (facts) that can be easily used in Theory Mapping e.g. facts can be easily downloaded and dragged into Maps to support theories.
  
• Personal profile: where you can bookmark your favourite theories, and get an email notification if the league table ranking of them changes.

Theory Mapping Software

f
The application can either be developed from an existing Argument Mapping software, or developed from scratch. The former is probably most efficient, since the changes required would not be that large.

The best Argument Mapping software currently available is bCisive, which already has a Hypothesis Testing mode that can be used. Although many argument visualisation softwares have the boxes and arrows to allow argument mapping, a key advantage of bCisive is that much of the construction and positioning of boxes and arrows is done automatically. This makes it much faster to use. For instance, if you want to insert a new reason box into the map, you just drag it in, and the layout of the map is automatically adjusted. Whereas if you want to use normal box and arrow software like Inspiration/Webspiration, Compendium, or cMapTools, you have to manually move all the other boxes out of the way first to create space. This is very time-consuming when you are dealing with a large map.

The key modifications to bCisive that are needed to make Theory Mapping sufficiently user-friendly to gain widespread use would be (in order of importance):

• Web-based: bCisive is currently in the process of being made web-based.
• Add a ‘Facts panel’: Agreed Facts and Background Facts would be inputted into this in manually (or transferred from a database on the website once this is developed), from which they can be easily dragged onto the map (similar to how different boxes can be dragged onto the map from the Building Panel). bCisive already has a ‘Fact’ box that can be used. When dragged onto a map, only the title of the fact would be shown inside the box (given how long some facts may be), but hovering over the map would show the full text.
• Notes: allow them to link to two boxes, so that inconsistencies between explanations and theories can be highlighted.
• Reasons that support more than argument: place these automatically in a layer below the other argument trees, with lines linking them to the arguments. Before this is developed, they can be placed below manually, with a reference to them placed in the top layer so that it is clear to the reader.
  
• Incoherence score: automatic calculation of the score based on Objections placed in the Theory Map and the Facts Panel.

Equally, a number of features in bCisive could be taken out in order to simplify the interface e.g. most of the content of the Building Panel and the Text Panel.
f
f

Theory Mapping Organisation and Funding

f
Organisation

An organization (or project within an existing organization) would need to be set up to run Theory Map Debates and Competitions.

Running Theory Map Debates and Competitions would consist of the following:

• Maintaining the website;
  
• Moderating Debates and Competitions according to a set of rules (this could be done by volunteers once the basic rules are established).
• Publicising Debates and Competitions.

Funding

The organisation would need money to ensure that the Theory Mapping tool is continually improved, while keeping it free for members of the public to use. Some potential sources of revenue could include:

• Seed funding: either private or public.
• Sponsorship of Debates: high profile Theory Map debates could be sponsored by either private or public entities, with the company charging a fee for this;
• Commission fee for Competitions: for Theory Map competitions, the company could charge a percentage of the prize money, payable upon successful completion of the competition;
• Online advertising: if Theory Mapping really takes off, it would have a high level of traffic;
• Public donations: since it is free to use, people may wish to contribute towards the cost of providing it.

f

Theory Mapping Pilot Test: Who Shot JFK?

f
Why choose the JFK assassination debate?

The debate about 'Who shot JFK?' may seem like a strange subject to choose to pilot test Theory Mapping, but actually they are a number of reasons why it would make an ideal test case:

• About a valid issue: both sides of the debate do have coherent arguments that should be respected, and many see the assassination as a turning point in American history. In particular. the JFK assassination has given rise to the modern form of conspiracy theories that some academics argue are based on sophisticated analyses of the forces at work in a modern globalised world. For instance, Peter Knight, a Senior Lecturer in American Studies at the University of Manchester, and author of the book The Kennedy Assassination writes that "It's arguable, therefore, that Kennedy assassination conspiracy theories are no longer necessarily the sign of crackpot, delusional thinking that borders on clinical paranoia...The public support for the findings of Kennedy assassination critics also suggests that conspiracy theories have moved from the fringe to the mainstream of American society" (p.98,99) Indeed, the vast majority of Americans believe in some form of conspiracy theory about the assassination.
• Current lack of progress: a number of commentators believe that the JFK assassination debate, although pioneering in its heyday, has since got stuck in a rut. As Knight argues, "There has been no steady convergence of opinion about the case, not only between the lone gunman and conspiracy theory camps, but within the latter position" (p.99)
• Heavy duty problem: the lack of progress may be traced to a number of fundamental problems. These problems are exactly what Theory Mapping is designed to solve, and so the debate will test the method to the full. The problems include: a) Inability to falsify theories by testing hypotheses i.e. it is not amenable to the scientific method (Theory Mapping does not require falsifiability); b) Involves a large number of facts that have been amassed by the different sides over the years, which is hard to collate: "...the problem is not that there is too little evidence to solve the case but too much data for anyone to master easily..." (Knight p.100) (Theory Mapping provides a way to systematically assess theories in terms of any number of facts); c) It involves different world views, since conspiracy theorists make different assumptions about the trustworthiness of government and hence treat the evidence collected by government agencies differently from those adhering to the ‘lone gunman theory’. "...it has produced a free-fall of suspicion that begins to doubt everything - even the fundamental ground rules of proof and evidence" (Knight p.100) (Theory Mapping allows different world views to have their own rules of proof and evidence, providing they are coherently justified); d) Distorted incentives and lack of quality control, leading to misinformation, back-biting and fraud (by having an objective test of the truth, Theory Mapping ensures that everyone's incentives are correctly aligned).
  
• High profile issue: the JFK assassination is a high-profile issue which many people feel passionately about. There is therefore likely to be a good supply of people willing to engage in the debate and overcome the initial costs of learning how to use the Theory Mapping method.

Implementation Plan

The plan to implement the pilot test is currently as follows:

1. Identify a software company and/or research institution that share the vision for Theory Mapping and are interested in sponsoring the first debate.
2. Find two high-profile JFK experts to conduct the first debate, one for the Lone Gunman Theory, and one for a version of conspiracy.
3. Run the first debate in private while recording all the content.
4. Evaluation of the private stage of the pilot test, with a report to partners on the debate's performance and lessons learnt.
5. If successful, secure seed funding to set up the organization, software and website. This would start with the content from the first JFK debate and open the debate up to the public, with marketing through the various JFK discussion boards.
6. Evaluation of the public stage of the pilot test, with a report to partners on performance and lessons learnt.
7. Modifications made as appropriate.
8. Full launch with debates and competitions on any subjects possible.
f

Objection: Too complicated

f
Objection:

Theory Mapping is simply too complicated to be able to command a widespread following, requiring significant upfront investment of time in agreeing on the facts, and learning and drawing up the Theory Maps.

Responses:

1. Importance of good software
The level of time and mental effort required can be significantly reduced through effective software. For instance, the software can have a function for entering facts, which allows them to be dragged and dropped onto each map. See Theory Mapping Software for a discussion of this.

2. Upfront costs decline over time
The level of effort required by a newbie would decline significantly over time for a particular topic, as a database of facts is drawn up and example Theory Maps exist to refer to.

3. Incentives outweigh costs
Having an objective test for who wins and loses, creates a strong incentive to participate or observe, that can outweigh the upfront investment costs.

4. Widespread uptake isn't necessary
Even if Theory Mapping does not command a widespread following, it still has the potential to revolutionise thinking and debate on a subject. This is because those who do participate are likely to be the opinion leaders, and the results will be widely publicized.

5. Argument Map literacy will improve over time
It has been pointed out by Simon Buckingham Shum (here) that with regard to argument visualization tools we need “a new literacy in being able to read and write in the new medium, and a new fluency in using these conversational tools in appropriate ways in different contexts.” As argument mapping becomes more widely used, so this literacy will develop.

6. Need for pilot testing
The only way to be sure of whether Theory Mapping is too complicated is to pilot test it!
f