Show Me Your Map by Timothy J. Harland A map-maker’s guide Piece 2 of 3 — Chapters 6–10 From the Author’s Portal · showyourmap.com Free for weary travelers. E UNUM PLURIBUM. ------------------------------------------------------------ CHAPTER 6 What If Both Perspectives Are Useful? Different perspectives do not have to cancel each other out. Sometimes they are answering different questions. Two Routes Imagine two people choosing a route across the same city. One chooses the fastest. The other chooses the safest. Their maps may produce different paths. That does not necessarily mean one misunderstood the city. One map gives more weight to time. The other gives more weight to risk. Ask which route is correct and the disagreement may look larger than it is. Ask: What does each route make visible? and the relationship changes. Different Representations A road map and a topographical map can describe the same place. One makes roads, intersections, and distances easy to see. The other reveals elevation, slopes, waterways, and terrain. Neither needs to contain everything the other contains. Their usefulness depends partly on what we are trying to do. If we are driving, the road map may be enough. If we are deciding where floodwater is likely to travel, the topographical map may suddenly matter much more. The area did not change. The representation did. Compare Before Choosing When perspectives differ, we do not always need to decide immediately which one wins. 12 First compare them. Where do they overlap? Where do they diverge? What does one reveal that the other does not? Sometimes one perspective will prove more useful. Sometimes both will. And sometimes comparing them reveals something neither contained alone. The picture is getting broader. But broader is not the same as complete. What are we still not seeing? 13 CHAPTER 7 What Are We Still Not Seeing? Adding perspectives can make a picture clearer. It can also make completeness easier to assume. The Intersection Imagine several people observing a busy intersection. One watches traffic flow. Another watches pedestrians. Another watches delivery access. Together, they see considerably more than any one observer would. None of them notices the timing of the traffic signal. More perspectives improved the picture. They did not finish it. Sometimes what is missing is another perspective. Sometimes it is a variable no one thought to observe. Sometimes it is something everyone assumed was already understood. A useful question is: What would have to be missing for our current picture to still be incomplete? Change the Observation Position Before adding another explanation, try changing how the system is being observed. Describe what can be observed before deciding what it means. Then ask whether the same picture appears from another position. Perfect independence is unlikely. Everyone brings experience, expectations, and a point of view. The useful question is not whether an observer has no lens. It is also: Which influences does this observer share with the current reconstruction, and which do they not? Sometimes changing the observer reduces a familiar assumption. 14 Sometimes changing the interface changes what becomes visible. Sometimes operating inside the environment reveals constraints that disappear when viewed from outside. The differences between those observations are themselves information. Sometimes the missing variable is not inside the system we are observing. It is in the way we are observing it. Look Outside the Current Map Then compare the reconstruction with something outside it. A record. A measurement. A prior example. Another observer. An external source does not automatically settle the question. It gives us another comparison state. It may support what we already think. It may contradict it. Or it may introduce a detail no one had considered important. That detail does not have to become meaningful immediately. Its significance may simply be undetermined. Do not decide that an observation is useful or useless before its usefulness can be measured. Preserve it long enough for later comparison to earn the classification. Eventually one of those additions may change more than the amount of information we have. It may change the meaning of information we already had. Does this change what we thought we knew? 15 CHAPTER 8 Does This Change What We Thought We Knew? New information does not always require a new explanation. Sometimes it fits easily into the one we already have. Sometimes it does not. The Repeating Failure Imagine a piece of equipment that has failed several times. Each time, the same component was replaced. The equipment returned to service. The history seems straightforward: component fails → component replaced → equipment works Then a technician notices something. The fault disappears when a connector is moved. That observation does not make the previous replacements imaginary. Those components were replaced. The equipment did return to service. But the new detail changes what those events might mean. Maybe the component was failing. Maybe something else was causing it to fail. Maybe replacing the component disturbed the connection and temporarily restored operation. The old information remains. Its relationship changes. Compare Before Replacing When a new observation does not fit, there is an understandable temptation to throw out the old explanation. That may be unnecessary. Instead ask: 16 What still fits? What no longer fits? What needs another look? The answer may require a complete reconstruction. Often it requires much less. A connector may move from incidental detail to probable contributor. A repair history may shift from evidence of repeated component failure to evidence of a recurring pathway. The representation changes without requiring everything inside it to change. Keep What Still Works Revision does not mean the previous representation was foolish. It was built from the information available at the time. New information changes what can reasonably be said now. Keep what still works. Revise what does not. What should we do with the new version? 17 CHAPTER 9 What Should We Do With the New Version? A revised representation may be more useful than the one that came before it. That does not mean every possible response follows from it. A better map is not the same thing as a decision. What Changes Now? Imagine a process that has been revised after a hidden dependency is discovered. The new representation is clearer. Now what? Change the procedure? Add a safeguard? Watch for recurrence? Do nothing yet? The answer depends on what the new information actually makes possible. Response Has a Cost Every response uses something. Time. Effort. Resources. Attention. Sometimes disruption. Sometimes new risk. Continuation has a cost too, so the question is not simply: What could we change? It is: What does it cost to continue, and what risk does continuation carry? A deviation does not require correction merely because it exists. The response should fit its significance, the cost and risk of continuation, and what can actually be changed. Sometimes the smallest useful change is enough. Sometimes doing nothing costs more. Sometimes the best next step is to monitor, tolerate, preserve, or leave the condition alone. 18 For This Situation A revised representation may change a priority. It may change where we look. It may change a process. Or it may leave the current action unchanged while increasing what we know. That is still useful. The purpose of the new map is not to force movement. It is to make the next decision better informed. And the decision that works here may not work somewhere else. How far does this version travel? 19 CHAPTER 10 How Far Does This Version Travel? A useful representation may work well in one situation and poorly in another. That does not necessarily make it wrong. It may simply have limits. Try It Somewhere Else Suppose a process works well at one location. You apply the same representation at another similar location. It works there too. Then you try somewhere less similar. The scale is different. The environment is different. The people are different. The constraints are different. Maybe the model still works. Maybe only part of it does. Maybe it breaks completely. Each result tells us something. Similar Is Not Identical Two situations do not have to be identical for the same representation to be useful. But they do need to be similar in the ways that matter. That gives us a better question than: Does this work everywhere? Ask: Are these situations similar enough for this representation to remain useful? Sometimes the whole model travels. Sometimes only part of it does. Sometimes none of it should. 20 Where Does It Break? Limits are information too. A model that works for ten people may not work for a thousand. A process that works with abundant resources may fail when those resources disappear. A decision that works under normal conditions may become inappropriate during an emergency. The breakpoint helps show which assumptions were carrying more weight than we realized. A representation does not need to be universal to be useful. It only needs to travel far enough for the question we are asking. But if some part keeps surviving the journey, that part deserves a closer look. What has to remain the same? 21