Causal Map

A causal map is any diagram that uses nodes (things that matter) and arrows (influences) to explain how one thing leads to another. It is a way of turning “stories about what causes what” into something a group can point at, argue with, and improve. In practice, causal maps sit on a spectrum from simple linear chains to messy networks, and different traditions give the same basic idea different names, depending on what they want the map to do.

# Causal mapping as the umbrella “Causal mapping” is the umbrella term. Under it you will find workshop methods, planning tools, and systems-thinking diagrams that all share the same core move: represent a situation as a directed graph of causal influence - ifm.eng.cam.ac.uk Some causal maps are mainly for negotiation and shared understanding. Others are meant to support formal modelling. Most sit somewhere in between.

# Problem Trees as hierarchical causal maps A Problem Tree in Project Cycle Management is a causal map with a strong preference for hierarchy. The aim is to organise a negative situation into cause-and-effect layers so that smaller, concrete problems can be traced upward into fewer, more abstract “top level” problem statements - wikis.ec.europa.eu This “tree discipline” is partly about clarity. By avoiding cycles, the group can see “roots” that drive multiple outcomes, and “peaks” that summarise the lived experience of many underlying causes.

# Objective Trees and Quest Trees An Objective Tree is the positive mirror of a Problem Tree: each negative condition is rewritten as a positive desired condition, and the hierarchy becomes a means-to-ends structure. If you prefer the word “quest” because it is easier to picture, you can treat Quest Tree as a human-friendly alias for Objective Tree - wikis.ec.europa.eu The important point is that the structure is still causal, just in “how we get to a better future” form.

# Causal Loop Diagrams as causal maps with feedback A Causal Loop Diagram (CLD) is also a causal map, but it makes one design choice explicit: feedback loops are expected, named, and used to explain behaviour over time - wikipedia.org CLDs usually add two extra pieces of meaning. Arrows are labelled with polarity (+ or −) to show whether variables change in the same direction or opposite directions. Loops are labelled as reinforcing or balancing to show whether the feedback amplifies change or counteracts it. So, a neat way to say it is: a CLD is causal mapping that is optimised for feedback thinking.

# The relationship in one mental picture If a Problem Tree is a “dig down to roots, climb up to peaks” map, a CLD is a “follow the arrows until you come back round” map. A Problem Tree is excellent when you need prioritisation and a shared hierarchy. A CLD is excellent when you need to explain persistence, oscillation, runaway growth, “fixes that fail,” and other dynamics created by feedback. In real projects you often do both: start with a Problem Tree to get agreement on the landscape, then convert the most important chains into a CLD to surface the vicious circles and virtuous circles hiding in the story.

# Where cognitive mapping and SODA fit In the strategy and decision-support world, “cognitive mapping” and “causal mapping” are often used for facilitated group sensemaking. One well-known approach is SODA (Strategic Options Development and Analysis), which builds maps from interviews and workshops and then uses them to support negotiation and option choices - ifm.eng.cam.ac.uk In that framing, Problem Trees and CLDs can be seen as specialised dialects of a broader causal-mapping craft.

# When you want the peaks to “emerge” cleanly If your goal is explicitly to produce clear tiers of abstraction, there are more formal structuring methods that turn relationship judgements into levels. One example is Interpretive Structural Modeling (ISM), which is designed to structure many related elements into a multi-level model - mdpi.com You do not need ISM to run a good PCM workshop, but it is useful language if you want to explain “how we turned a messy graph into levels.”