Features
See how the argument is connected
Every recommendation is a structure: a conclusion held up by reasons, which are held up by evidence, which is sometimes missing. Tracing that structure, even roughly, shows where one weak claim is carrying far more than its share. That is the purpose of a decision graph.
The shape of an argument
In The Uses of Argument, published in 1958, the philosopher Stephen Toulmin broke practical arguments into parts: the claim, the grounds offered for it, the warrant linking the two, the backing for that warrant, the qualifier that limits it, and the rebuttal that would defeat it. Most business memos state the claim and the grounds and leave the rest unspoken. A decision graph brings the unspoken parts into view.
Look for the load-bearing node
Once claims are connected to the scores they support, a pattern often appears: a single assumption sits beneath several scores at once. If it fails, the leading option falls on three criteria together. That claim deserves checking before anything else, and it is often the one nobody thought to question.
The opposite pattern matters too. An option whose scores rest on several independent, supported claims is sturdier than its total alone suggests.
How the connections appear in Argumont
In a Decision Room, the arguments are broken into claims, and Inspector marks each one supported, partly supported, unverified, or contradicted. The Judge decides what survived, and the Verdict states what the leading option rests on, what remains uncertain, and the conditions that would change it. Those statements are the graph in words: they name the load-bearing claims and what follows if they fail.
The Decision Passport keeps the question, options, evidence, Verdict, and Human Decision together, so a reviewer months later can trace the reasoning without reading the whole debate again.
Reading it before you decide
Before committing, put three questions to the record: which claim, if false, would change the answer; whether that claim is supported or only assumed; and what it would cost to check. If checking is cheap, check first, so the choice is made with the weakest link in view.
One assumption under three scores
A team compares building a scheduling tool with buying one. The build option scores well on cost, fit, and control. Traced back, all three scores rest on one claim: that two engineers can deliver the tool in ten weeks. Inspector marks it unverified, since the team has never built anything similar. If the build takes twenty weeks instead, its cost rises by $70,000 at $3,500 per engineer week, and its fit score stops mattering because the busy season will have passed. The Verdict favors buying, and names the one fact that would reverse it: a comparable internal project delivered on time.
Questions
- Do I need to draw the graph myself?
- No. The room breaks the arguments into claims, and the Verdict states what the leading option rests on. Sketching it on paper is still a good habit for any choice made outside a room.
- Who reads the connections later?
- Anyone reviewing the decision: a manager, an auditor, or you a year from now. The Passport keeps the record, and a saved version cannot be overwritten.
- What if the load-bearing claim cannot be checked?
- Then the Verdict says so, and the uncertainty becomes part of the record. Sometimes the right move is to wait for the fact, and sometimes to choose the option that does not depend on it.
Bring Argumont a Decision worth challenging.
Start with a paid 60 day pilot on your own real Decisions.