… I made a blunder, my dear Watson——which is, I am afraid, a more common occurrence than anyone would think who only knew me through your memoirs.
—— Silver Blaze
I confess that I have been as blind as a mole, but it is better to learn wisdom late than never to learn it at all.
—— The Man with the Twisted Lip
Those who have experience in following clues will know that in this kind of activity we spend a lot of time correcting our own mistakes. We have to correct our own mistakes because there is no one else to do it for us. We follow clues when there is no one there to tell us the things we want to know. If there were, that same person would also be able to tell us where we have made mistakes and how to correct them. But there is no such person. Any mistakes we make, we ourselves will have to find them and correct them; we cannot rely on anyone else, any one that is already in the know.
But how can this be done? How do we correct our own mistakes when following clues? This frequently puzzles bystanders. Suppose a crime has been committed and we are still trying to find out what happened. Suppose in the process, unbeknownst to ourselves, we have made mistakes A, B, and C. Since we do not yet know what happened, how do we know we have made mistakes A, B, and C?
No one likes to make mistakes. Even less do we like to admit we have made them when we have. Indeed, some people believe they don't ever have to admit they have made mistakes. They have found the Truth, they believe, and they can defend it come what may (such is the power of words in their view). But in following clues we learn very quickly that we cannot do this. In following clues even the proudest person will have to acknowledge over and over again that she has made mistakes. Sherlock Holmes himself is a good example. Is there a person more confident of his own talents? Yet Sherlock Holmes is more than willing to admit that he can make, and has frequently made, mistakes (see the opening quotes to this chapter).
Why should the clue-following process have this effect on people? What is it about this process that makes us willing to admit to mistakes we ourselves have made? Remember we follow clues when the things we want to know are not yet known. They are not yet known to us, they are not yet known to our audience. In such a situation if we deny we have made mistakes who will be any the wiser?
But of course we do not do this. In practice, when we are following clues we, like Sherlock Holmes, will be more than willing to admit to mistakes we have made. Why? What compels us? And in practice how do we correct mistakes? How, when there is no one there to guide us, no one already in the know?
The way we correct mistakes when following clues is an intriguing subject. We take it up in this chapter.
We have pointed out in earlier chapters that if we want to be successful when following clues we must not mind making mistakes. Mistakes are common when following clues. There are a number of reasons, the most important of which is the simple fact that the things we want to know are hidden. If they were not; if they were there for everybody to see; mistakes would be much fewer in number. As the saying goes, when we are following clues we are working in the dark. People who work in the dark are bound to bump into things, not now and then, but frequently.
It is true in this process we do not simply grope in the dark; instead, we follow clues. But clues do not carry labels announcing to the world they are clues. Clues have to be detected. How do we detect them? It depends on the theories and assumptions we have advanced, theories and assumptions that could be wrong, for which reason it quite often happens in practice that something we thought was a clue turns out not to be.
Also, clues have to be interpreted. It is well known that often there is more than one interpretation to a clue. Sometimes the number of possible interpretations can be huge. Since not all interpretations of a clue can be right, before we hit upon the right interpretation we frequently will have gone through a large number of wrong interpretations.
The clue-following process is complex. Error could enter at many points. No one engaged in this process should be surprised to find that despite all the care they have taken they nevertheless have made mistakes.
Mistakes in an investigation fall into kinds: major and minor. Minor mistakes are easier to correct; they are simply those that can be corrected as the investigation advances. Major mistakes causes an investigation to come to a halt. Since they prevent an investigation from advancing they cannot be corrected in the same way as minor mistakes. The correction of major mistakes is much harder. We deal with major mistakes in the next two sections. In this we confine ourselves to minor mistakes.
We use the terms 'major' and 'minor' here in a theseological context, that is, in the context of following clues. A minor mistake in this sense could have serious consequences. For example, it could lead to a declaration of war and thus, in a political sense, not minor at all.
Why is it that some mistakes can be corrected as the investigation advances? Why, when we are still looking?
The answer is simple. As an investigation advances we find out more. Once we know more context will tell us where the mistakes are and how to correct them. Take our SBR cryptogram for example.
SBR SBCTU DBCKERVS FCGG WTTCXR SFH FRRJD YTHE SHUWI
Suppose as a first step we decipher FCGG as WALL. After this first step we will not know if we have made any mistakes. But if we continue it will not take us long to discover that FCGG cannot be WALL; it has to be WILL; WALL cannot fit in with our later decipherments; if we insist on WALL it will stand out like a sore thumb. From this we see correcting minor mistakes is in fact not all that different from correcting typographical mistakes. When proof-reading we do not look at words one by one and ask of each whether it is a mistake. We read sentences and paragraphs and from context detect where the mistakes are so that we can correct them. When following clues we do the same thing; that is, we also depend on context. Clues do not reveal to us the truth in one single step; they do so gradually. This is to say, as the investigation advances we will know more and more. When this happens context will tell us where the mistakes are; at that point it will also tell us how to correct them.
Compared to major mistakes the correction of minor mistakes when following clues is simple. Ordinarily most people will be more than happy to do what is required. After all, we want to know. We want to know what the clues are hiding. In this process removing mistakes is just one of the necessary steps; indeed, part of the routine.
In an investigation we frequently make mistakes. Unfortunately for us, not all the mistakes we make are minor. If they were, the life of the investigator would have been much simpler. But since they are not; since major mistakes can also occur; the life of the investigator is much more difficult.
Major mistakes lead to impasse; their presence in an investigation makes progress in the same investigation impossible. Major errors are hard to locate and also hard to correct.
There are two kinds of major errors in an investigation: single and multiple. This is to say, impasse in an investigation could be caused by a single major mistake or by more than one. We do not go about correcting these two kinds of major errors the same way. In this section we explain how to locate and correct single major errors; In the next, how to locate and correct multiple major errors.
To locate a single major error, the common practice is to retrace our steps. We examine our latest step first. If we do not find the error there, we go back to the one before, and the one before, and so on. Since major errors cause impasse usually soon after they have been made this kind of retracing is a sound practice.
How do we determine whether we have located the error causing the impasse? How do we go about finding a single major error?
Here we come to an intriguing question. We have to remember, if major errors were easy to detect, and thus also easy to correct, we would not have called them major in the first place. But if they are not easy to detect, how are we to locate them? And if we cannot locate them, how can we correct them?
What do we do in practice when we suspect a single major error is causing impasse? We have said we retrace our steps, but how do we know which step is the mistake causing the impasse?
If we are put into an actual situation we will find this is what we will end up in doing. We try to locate and correct the mistake at the same time, through trial and error.
Let me explain. Suppose we suspect the single major error is in step A. We re-do step A, making sure that the new step A is different from the old step A. If after this re-doing the investigation is able to advance we know the old step A is the error and the new step A, its correction.
What if the investigation is still unable to advance after the re-doing?
In this case one possibility is that old step A is not the mistake, in which case we will have to try a different step. Of course there could be other possibilities besides this but I will leave to readers themselves to work out what they are. In practice if we succeed in finding and correcting the single major error we frequently will have done so after going through this tedious procedure many times.
Notice that this procedure we have just outlined will only work if there is only one single major error. This procedure will not work if there is more than one major mistake. As we have mentioned earlier the correction of multiple major errors is different from the correction of single major errors.
Why is it that multiple major errors cannot be corrected the same way as single major errors? The answer to this question is simple. When we have a single major error, we know it has been corrected when we are able to advance again. But when we have multiple major errors, even if we should succeed in correcting most of them, so long as there is one left uncorrected, advance is still impossible. To advance again, we have to correct all the mistakes at the same time and this is difficult to do.
Because multiple major errors are difficult to locate, when their presence is suspected, it is sometime simpler to ‘start all over again’, or, at least, to re-do that part of the investigation in which the errors are suspected to have occurred. The hope here is, by re-doing the whole investigation, or the relevant part, we will not make the same mistakes a second time.
When dealing with multiple major errors this way——that is, by redoing the whole investigation, or a part of it, it is useful to call in new blood. By keeping them in the dark as to what has happened in previous attempts, the chance is greater that the same mistakes will not be made.
Notice that in trying to re-invigorate an investigation by re-doing it, it is not necessary that we know exactly what mistakes were made in the past, and how. Whatever they were, and however they were made, so long as we are able to advance beyond the point at which the investigation was stalled, and to keep on advancing, these mistakes must have been corrected. The pinpointing of mistakes made in the past, and explaining how these mistakes were made, are not always easy tasks even from the vantage point of a resuscitated investigation.
Since the correction of multiple major errors differs from the correction of single major errors, it is important that, when troubleshooting a stalled investigation, we do not take for granted that we are dealing only with a single major error. We may suspect the presence of a single error first, but if this suspicion produces no results, we should go on to suspect the presence of multiple errors. If we take for granted that we are dealing with a single major error when in fact there is more than one, we could be making things worse rather than better. For example, suppose there are two major errors, and we have taken for granted there is only one. Suppose one of these errors lies with step A, and the other with step B. Suppose we first suspect A to be the location of the single major error. So we re-do A, hoping that after re-doing it, we should be able to advance again. But we cannot, since there is still step B. From this, we conclude, erroneously, there is nothing wrong with A. We then re-do B, hoping that after re-doing it, we should be able to advance again. But again we cannot. From this we conclude, again erroneously, there is nothing wrong with step B. So still thinking that we are dealing only with a single major error, we look for it elsewhere——while convinced that A and B are both correct!
When major mistakes are made in an investigation progress will become impossible. For the investigation to advance again these mistakes will have to be located and corrected. This is not an easy task, as we have explained, and can be time-consuming. To save time; to avoid being held up by major mistakes; we sometimes carry out independent, parallel investigations. For example, we sometimes have two teams of detectives working on the same case, or two teams of scientists working on the same problem, with little or no communication between them. This way, if one team is held up by major mistakes, the other team hopefully could still continue.
When we set up independent, parallel investigations we prefer that the groups of investigators do not communicate with each other (this is why we call them independent), or if communication cannot be totally prevented, that it be kept to a minimum. We do this to prevent these groups from making the same mistakes. People who constantly communicate with each other have a tendency to look at things roughly the same way, so that if one makes a mistake the will make the same mistake as well.
Supporting a number of independent, parallel investigations appears on the surface to be wasteful of resources: it costs more to support two teams of detectives than just one. But when we are dealing with difficult investigations, investigations in which it is easy be held up by major mistakes, in the long run it may turn out to be more economical.1
When we carry out investigations we have to make assumptions about those things we are investigating. In the investigation of nature (science), we have to make assumptions as to what nature is like. Now when two teams of investigators carry out independent, parallel investigations, their assumptions could be different. But different assumptions mean different ways of looking at the world, or different ways of looking at that part of the world with which the investigations are concerned. Since independent, parallel investigations are desirable (because economical in the long run), if we want our knowledge to expand; if we want to prevent Dark Ages from setting in, that is, periods in which no progress is made; we should not only allow, but encourage, individuals and groups of individuals to look at the world in different ways.
When Sherlock Holmes discovers he has made mistakes he is quite willing to admit to them. So would most of us when following clues. Why? Why would we do this instead of insisting we have been right all along? As we will see, this is not an idle question but has important consequences. Also, it is a question we can now answer now that we have come to the end of this chapter.
There are two kinds of mistakes when following clues: major and minor. What is our reason for admitting to them when we have made them?
Let us take minor mistakes first. Minor mistakes are those that we can correct as the investigation advances. This is to say, this is the kind of mistake that becomes obvious as more is known. So, why do we admit to them? Why do we not refuse instead? To this question the answer can only be, this kind of mistake is so obvious; it will take an extremely strong-willed person to refuse to admit to obvious mistakes. After all, what are we trying to do when following clues? We are trying to know; and the knowledge we have gained now tells us we have made an mistake. We made the mistake when we did not know, or did not know enough. Now that we know more it is only natural that we should correct the mistake we made earlier.
So much for minor errors. What about major errors? Why do we admit to them when we discover we have made them?
What are major error? They are those that cause impasse.
How do we know we have corrected a major error? We know when the investigation is able to advance again.
What happens if we refuse to admit to major errors? Here we discover the reason why most people would not refute to admit to major errors. For to refuse would be tantamount to giving up the search! We follow clues because there are things we want to know. If we ever succeed in finding the things we want to know the success can only come as the result of a long search. Major errors interrupt this search. If we do not admit to them the search cannot resume.
If we understand the clue-following process we will find there are good reasons why those who are good at it are willing to admit to mistakes they have made. We can put this same point in the form of a question. If you are unwilling to admit to mistakes can you ever become Sherlock Holmes?
1Independent, parallel investigations are common in science.