Eight to Late

Sensemaking and Analytics for Organizations

Why visual representations of reasoning are more effective than prose

with 11 comments

At work I’m often saddled with tasks that involve writing business documents such as project proposals, business cases, technology evaluations etc. Typically these are aimed at conveying positions or ideas to a specific audience. For example, a business case might detail the rationale behind (and hence the justification for) a project to executive management. The hardest part of composing these documents is the flow: how ideas are introduced, explained and transitioned one after the other.  (Incidentally, writing business documents is composition – like music or art – let no one tell you otherwise.) The organisation of a document has to be carefully thought through because it is hard to convey complex, interconnected ideas in writing. Anyone who has laboured through a piece of reasoning in prose form is familiar with this problem.  In an earlier post I discussed, via example, the utility of issue mapping – a visual representation of reasoning – in clarifying complex issues that are presented in writing. In this post I explore reasons why issue maps – or other visual representations of reasoning – are superior to prose when it comes to conveying ideas.  Towards the end, I also highlight some potential uses of visual representations in project management.

The basic problem with prose is that the relationships between ideas are not immediately evident. In a paper entitled Enhancing Deliberation Through Computer Supported Argument Mapping Tim van Gelder states,

Extracting the structure of evidential relationships from reasoning as typically presented in prose is very difficult and most of the time we do it badly. This can be easily illustrated, in a kind of exercise we have done informally many times in workshops. Take any group of people sufficiently trained in reasoning and argument mapping that they are quite able to create argument maps to make explicit whatever reasoning they have in mind. Now give them a sample of good argumentative prose, such as a well-argued opinion piece from the newspaper. Ask them to figure out what the reasoning is, and to re-present it in an argument map. This usually takes about 20-30 minutes, during which time you can enjoy watching the participants strike various Rodinesque postures of intense concentration, wipe their sweaty palms, etc.. Then compare the resulting argument maps. You’ll find that you have as many different argument maps as there are people doing the exercise; in many cases the maps will be wildly different…

Although the paper is talks about argument mapping, the discussion applies just as well to issue mapping. He goes on to say,

Take any group of people sufficiently trained to be able to be read argument maps. (This training usually takes not more than a few minutes.) Present them with an argument map, and ask them to identify the reasoning presented in the map, and represent it in whatever form they like (map, prose, point-form etc.). This is a very simple task and usually takes almost no time; indeed, it is so trivial that the hard part is getting the participants to go through the motions when no intellectual challenge is involved. Ask them questions designed to elicit the extent to which they have correctly identified the structure of the reasoning presented by the map (e.g., how many distinct reasons are presented for the main conclusion?). You’ll find that they all understand exactly what the reasoning is, and ipso facto all have the same sense of the reasoning…

The point is simple: visual representations of reasoning are designed to present reasoning; prose isn’t.

Van Gelder then asks the question: why are visual representations better than prose? In answer, he makes the following points:

1. Prose has to be interpreted: As prose is not expressly designed to represent reasoning, readers have to decode relationships and connections between ideas. The choices they make will depend on individual interpretations of the meaning of words used and the grammatical structure of the piece. These interpretations will in turn depend on facility with the language, vocabulary etc. In contrast, visual notations such as IBIS (used in issue mapping) have few elements and very simple grammars. Simplicity slays ambiguity.

2. Prose neglects representational resources: Prose is a stream of words – it does not use other visual elements such as colour, shape, position or any graphical structures (trees, nodes, connectors). The brain processes comprehends such elements – and the visually apparent relationships between them – much faster than it can interpret prose. Hence the structured use of these can lead to faster comprehension. Van Gelder also adds that in the case of prose:

…Helpful authors (of prose) will assist readers in the difficult process of interpretation by providing verbal cues (for example, logical indicators such as “therefore”), although it is quite astonishing how frugal most authors are in providing such cues….

This is true, and I’d add that writers – particularly those who write analytical pieces – tend to be frugal because they are taught to be so.

3. Prose is sequential, arguments aren’t: Reasoning presented in written form flows linearly – i.e. concepts and ideas appear in sequence. A point that’s made on one page may be related to something that comes up five pages later, but the connections will not be immediately apparent unless the author specifically draws attention to it. Jeff Conklin makes the same point about conversations in this presentation: conversations are linear, one comment follows the other; however the issues that come up in a conversation are typically related in a non-linear way. Visual maps of reasoning expose these non-linear connections between issues in a very apparent and easy-to-follow way. See this map of a prose piece or this map of a conversation, for example.

4. Metaphors cannot be visually displayed in prose: According to the linguist / philosopher George Lakoff, metaphors are central to human understanding. Further, metaphors are grounded in our physical experience because our brains take input from the rest of our bodies (see this interview with Lakoff for more). For this reason, most of the metaphors we use to express reasoning relate to physical experience and sensation: strength or weakness of an argument, support for a position, weight of an idea, external pressure etc. Van Gelder claims that visual representations can depict these metaphors in a more natural way: for instance, in IBIS maps, cons are coloured red (Stop) whilst pros are green (Go).

The above points are taken from van Gelder’s paper, but I can think of a few more:

5. Visual representations have less ambiguity: All visual representations of reasoning that I’ve come across –  mind maps, issue maps, argument maps etc – excel at displaying relationships between ideas in an unambiguous manner.  One reason for this is that visual representations generally have a limited syntax and grammar,  as a consequence of which a given relationship can be expressed in only a small number of ways (usually one!).  Hence there is  little or no ambiguity in depicting or interpreting relationships in a visual representation. This is not the case with prose, where much depends on the skill and vocabulary of the writer and reader.

6. Visual representations can present reasoning “at a glance”: A complex argument which takes up several pages of prose can often be captured in a single page using visual notations. Such visual representations -if properly constructed – are also more intuitive than the corresponding prose representation. See my post entitled, Beyond words: visualising arguments using issue maps for an example.

7. Visual representations can augment organisational memory A well structured archive of knowledge maps is so much more comprehensible than reams of documentation. Here are a couple of reasons why:

  1. Maps, unlike written documents, can capture the essence of a discussion minus all the conversational chaff.
  2. Maps can be structured to show the logical relationships or interconnections between multiple documents: i.e. as maps of organisational knowledge. Much like geographical maps, these can help knowledge workers navigate their way through vast tracts of organisational knowledge

See my post on knowledge capture using issue maps for much more on this.

8. Visual representations can catalyse knowledge creation: Visual representations, when used collaboratively, can catalyse the creation of knowledge. This is the basis of the technique of dialogue mapping – see this post for a simple example of dialogue mapping using IBIS. A visual representation serves as a focal point that captures a group’s collective reasoning and understanding of an issue as it evolves. Even more, the use of such representations in design discussions can foster creativity in much the same way as in art. In fact,  Al Selvin refers to Compendium – the tool and methodology used in dialogue mapping –  as an enabler of  Knowledge Art.  I’m currently reading some of Selvin’s writings on this, and will soon write a post summarising some of his ideas. Stay tuned.

If  you’re a project manager and have read this far, perhaps you’re wondering how this stuff might be relevant to your day-to-day work.  Well, in a couple of ways at least:

  1. Visual representations can serve as a succinct project memory.  A practical way to start is by using issue mapping to capture meeting minutes.  One can also use it to summarise meetings after they have taken place, but doing it in real-time is better because one can seek clarifications on the spot. Better still – project the map on to a screen and become a dialogue mapper. 
  2. Maps can be used in real-time to  to facilitate collaborative design – or create knowledge art –  as discussed in point 8 above.

To summarise, then: visual representations of reasoning are more effective than prose because they are better at capturing the nonlinear structure of arguments; easier to interpret; leverage visual metaphors; depict relationships effectively and present arguments in a succinct yet intuitively appealing way. Above all, visual representations can facilitate collaborative creativity – something that prose simply cannot do.

11 Responses

Subscribe to comments with RSS.

  1. […] for software development (Fred Brooks) entwickelt. In einem zweiten Posting hat er gestern seine Diskussion für die Visualisierung einer Diskussion und gegen Prosa […]

    Like

  2. […] a recent post,  Andreas Heilwagen  suggests that the arguments presented in my article entitled Why visual representations of reasoning are more effective than prose  might be better made through visual representation.  I reckon he is absolutely right. So here it […]

    Like

  3. Neat article Kailash. I’m still confused about the difference between issue maps and mind maps. We use mind maps a lot for ideas/project management etc. Where does issue maps fit in?

    Like

    Prakash

    May 27, 2009 at 12:31 pm

  4. Prakash,

    As I see it, mind maps are relatively unstructured – there is no grammar associated with how one develops a mind map starting from a central idea. Issue maps, on the other hand, follow the IBIS notation and grammar (described in
    this post
    ). There are three elements – issues (or questions), ideas and arguments (pros and cons) which can be connected only in ways specified by the grammar. I have summarised the grammar in figure 2 of this post.

    IBIS has just enough structure to make maps easy to follow (compared to mind maps). Further, the structure of IBIS makes it easy to map reasoning and argumentation – questions, answers and pros and cons are clearly marked. It is much harder to make these distinctions on a mind map.

    Hope this helps, and thanks for the appreciative comments.

    Regards,

    Kailash.

    Like

    K

    May 27, 2009 at 9:41 pm

  5. Well explained Kailash. Got it now. I have downloaded Compendium and am trying my hands at it. Thanks to your posts

    Like

    Prakash

    May 28, 2009 at 2:56 am

  6. […] respond to this question aren’t immediately obvious. This isn’t unusual:  as I’ve mentioned elsewhere, points can only be made sequentially – one after the other – when expressed in prose.  In […]

    Like

  7. […] (discussion) at any time can be inferred at a glance (unlike the case in written records). See this post for more on the advantages of visual documentation over […]

    Like

  8. […] behind decisions in a visual, easy-to-understand way.  The visual aspect is important – see this post for more on why visual representations of reasoning are more effective than […]

    Like

  9. […] for software development (Fred Brooks) entwickelt. In einem zweiten Posting hat er gestern seine Diskussion für die Visualisierung einer Diskussion und gegen Prosa […]

    Like

  10. […] Eight to Late on “Why visual representations of reasoning are more effective than prose”. I actually disagree with this one a bit: I’m suspicious of any argument that can’t (or won’t) be boiled down to simple language. But visual aids absolutely do help enhance an argument. […]

    Like

  11. […] (discussion) at any time can be inferred at a glance (unlike the case in written records). See this post for more on the advantages of visual documentation over […]

    Like


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: