Frequently, we decide on a goal, and then we are ineffective in working towards this goal, due to factors wholly within our control. Failure modes include giving up, losing interest, procrastination, akrasia, and <a href=”http://lesswrong.com/lw/2p1/a_failure_to_evaluate_returnontime_fallacy/“>failure to evaluate return on time</a>. In all these cases it seems that if our motivation were higher, the problem would not exist. Call the problem of finding the motivation to effectively pursue one’s goals, <i>the problem of motivation</i>. This is a common failure of instrumental rationality which has been discussed from numerous different angles on LessWrong.
I wish to introduce another approach to the problem of motivation, which to my knowledge has not yet been discussed on LessWrong. This approach is summarized in the following paragraph:
<blockquote> We do not know what we value. Therefore, we choose goals that are not in harmony with our values. The problem of motivation is often caused by our goals not being in harmony with our values. Therefore, many cases of the problem of motivation can be solved by discovering what you value, and carrying out goals that conform to your values. </blockquote>
I will begin by making clear the distinction between goals and values. A goal is time-bound, contingent, and concrete. A value is permanent, unconditional, and abstract. A few goals are:
<ul> <li>Become a comedian.</li> <li>Bring about the Singularity.</li> <li>Graduate college.</li> </ul>
I argue that goals are always in service of values. If we look at <a href=”http://lesswrong.com/lw/sm/the_meaning_of_right/“>the terminal values in our decision tree</a>, they will be big, mysterious abstractions such as these. Why would I think this? Why not have concrete things as the terminal values of our decision tree?
Because it’s simpler. Consider this example. I can have a terminal value, “pleasure,” and then sub-goals to this terminal value: let’s say, “drinking,” “smoking,” and “eating.” On the other hand, I can have “drinking,” “smoking,” and “eating” all as terminal values, without the node “pleasure” anywhere in sight.
The latter seems rather un-parsimonious. Why have multiple specifics in my utility function, when I can have one generality that replaces them all? Scientific theories are developed by subsuming specifics in generalities. Also, I think that the case in the previous paragraph is a fairly representative example; that for every specific desirable, we can name a more general desirable, or a set of more general desirables, which it serves.
So it seems to me that the simplest theories of value are those which have mysterious abstractions as the terminal values in the decision tree. These mysterious abstractions are what I have called “values.” Our goals, in turn, are non-terminal nodes which point towards our values.
In order to achieve our values, we must figure out how to achieve them, and then do so. We must choose our goals, and carry them out. But, there is another step in the process: <i>we must figure out what we value</i>. It is this step which I wish to examine here.
I make the following claims:
<ul> <li>Every person has a relatively stable set of values.</li> <li>No person has a perfect understanding of what they value. There is a disconnect between what we value, and what we understand ourselves as valuing.</li> <li>It is possible, through introspection and analysis of experience, to discover more perfectly what one values.</li> <li>Some instances of the problem of motivation are caused by pursuing a goal which is not sufficiently in harmony with one’s values. In these cases the solution is to understand the nature of the disharmony, and change the goal, or the approach to the goal, to eliminate the disharmony. This will probably involve making new discoveries about one’s values.</li> <li>Pursuing a goal which is strongly in harmony with one’s values does not require effort or willpower; it comes naturally.</li> </ul>
Unfortunately, I do not have any scientific evidence to support these claims. Also, the usual qualifiers to self-help advice apply. I will conclude the post by giving several examples of how these principles have been played out in my life.
I spend a lot of time on creative projects. For years I struggled with the problem of not being able to finish anything I started. I now finish most of the things I start. This is probably due to many factors. One factor, I believe, is that I now put more careful attention into making sure that the message I have for each creative effort strongly embodies what I value. I concentrate more of what I value into each piece, and this is possible because I have a better understanding than I used to of what I value.
Most of this has to do with the detailed content of the pieces. It also has to do with the media I use. I used to write music, but I no longer do so. Now I write a lot of philosophy. This is better for me because the greatest value which music embodies is beauty, and the greatest value which philosophy embodies is truth. I value both truth and beauty, but I value truth more than I value beauty, and so philosophy is more gratifying for me.
The biggest mistake of my life so far was my attitude in a relationship with a certain girl. I spent a lot of time trying to “fix” her or “improve” her, mostly by giving her advice or knowledge about herself and her personality. The information that I gave her about herself, while true, hurt her self-esteem, and this weakened our relationship, ultimately contributing to our breaking up.
What I should have done, as I realized in retrospect, was to love and accept her for who she was, rather than trying to fix or improve her. This experience taught me the value of love. Love is something I value more than I ever knew. It has proven difficult to propagate the belief “I value love” throughout my mind, so that I often forget it. Therefore I suspect that I value love even more than I know.
(This might seem like a paradox, but it is not. When I say that “I value love more than I know,” I mean simply that I don’t feel, continuously and at a gut level, the value that I believe that love really has. This is similar to <a href=”http://lesswrong.com/lw/1l/the_mystery_of_the_haunted_rationalist/“>the rationalist who is afraid of ghosts</a>. He knows intellectually that ghosts don’t exist, but he hasn’t managed to convince the rest of himself of this.)
I value mystical experiences because they efficiently serve a significant number of my values. They involve intense positive emotions and aesthetic ecstasies, serving my values of happiness and beauty. They often give me new ideas, serving my value of truth, and indirectly serving my value of self-expression by providing me with things to write about. (I came up with the idea for this post while meditating this morning.)
For a long time I did not have a clear understanding of what I was trying to get out of mysticism. At times, therefore, it became a <a href=”http://lesswrong.com/lw/le/lost_purposes/“>lost purpose</a>, which did not get me anywhere.
I now practice mysticism with a clearer idea of what I am trying to accomplish. This has made my practices more effective. Mysticism involves subtle manipulations of one’s mental processes. Rather than engaging in mental processes which are simply “what you’re supposed to do in mysticism,” I now engage in mental processes which I know will lead to the results I want. This has been made possible by connecting the processes with the values they serve.
It is not only true that my mysticism has become more effective for me by connecting it with an understanding of my values; it is also true that it has become easier. Previously, I would exercise a significant amount of willpower to engage in mystical practices for an hour or two a day. Now, I meditate for three hours a day without any exercise of willpower. It doesn’t require willpower because I simply <i>want</i> to do it. I want to do it because it serves my values.
I wish to analyze in detail a particular way in which I previously went wrong with mysticism. I used to meditate in a half-lotus position. Because I was born with low hip flexibility, this position put a lot of stress on my body. I was initially in a lot of pain from this position, but I continued using it. Eventually the pain went away. Further down the line, I gave myself a knee injury from meditating in this position, which probably would have required surgery if I had continued to do this.
There are a few things that I learned from this incident. I was meditating in half-lotus to conform to my image of “spiritual purity,” based on cultural memes which I had absorbed. My knee injury made me realize that I had been on the wrong path in this respect, and I saw that I was wrong in valuing “spiritual purity.” It was not part of my value system, but something that had been imposed from outside.
My knee injury also made me realize that I had been wrong in putting my body through all of that unnecessary stress in the service of mysticism. This was one of the data points which allowed me to discover the value of self-love: of being nice to oneself.
Another data point which allowed me to discover the value of self-love came in editing the book I am working on. Editing this book has been quite painful, because I am so critical of it. Reading it makes me disgusted by it, and then I am disgusted with myself.
For example, about a week ago I came across two paragraphs which contained some statements that I felt to be wrong. I was so disgusted by these wrong statements that I could not bear to look at them. I had to spend days psyching myself up to the point that I could look at these paragraphs for long enough to delete them.
The basic problem here is that I am hyper-critical of my writing, and in criticizing my writing I am criticizing myself. This makes the editing process a terrible stressor on my self-esteem. There is a clear message about my values: I value self-love, but do not love myself enough. If I was nicer to myself, then I could edit painlessly, and I desire to edit painlessly.
A final example will illustrate once more the connection between values and the problem of motivation. I am a psychology major and a philosophy minor. I got the psychology major at the beginning of college, and the philosophy minor two and a half years later. I enjoy my philosophy classes, and find my psychology classes tedious. It is therefore easier to be motivated to do well in my philosophy classes than in my psychology classes. I chose philosophy at a time when I better understood my values than I did at the time when I chose psychology. If I had this knowledge of my values when I started college, I could have chosen a major that I would enjoy more and do better in.
I hope that the reader has found these examples illustrative. I have attempted to throw some light on the process of discovering one’s values and revising one’s goals to conform to these values. It seems to me that this is a basic tool of instrumental rationality. One cannot achieve one’s values without knowing one’s values. Furthermore, one cannot sustainably pursue anything other than one’s values, and many cases of the problem of motivation can be explained by this fact.
Knowing what you want is a prerequisite to getting what you want
Frequently, we decide on a goal, and then we are ineffective in working towards this goal, due to factors wholly within our control. Failure modes include giving up, losing interest, procrastination, akrasia, and <a href=”http://lesswrong.com/lw/2p1/a_failure_to_evaluate_returnontime_fallacy/“>failure to evaluate return on time</a>. In all these cases it seems that if our motivation were higher, the problem would not exist. Call the problem of finding the motivation to effectively pursue one’s goals, <i>the problem of motivation</i>. This is a common failure of instrumental rationality which has been discussed from numerous different angles on LessWrong.
I wish to introduce another approach to the problem of motivation, which to my knowledge has not yet been discussed on LessWrong. This approach is summarized in the following paragraph:
<blockquote>
We do not know what we value. Therefore, we choose goals that are not in harmony with our values. The problem of motivation is often caused by our goals not being in harmony with our values. Therefore, many cases of the problem of motivation can be solved by discovering what you value, and carrying out goals that conform to your values.
</blockquote>
I will begin by making clear the distinction between goals and values. A goal is time-bound, contingent, and concrete. A value is permanent, unconditional, and abstract. A few goals are:
<ul>
<li>Become a comedian.</li>
<li>Bring about the Singularity.</li>
<li>Graduate college.</li>
</ul>
A few values are:
<ul>
<li>Humor</li>
<li>Transcendence</li>
<li>Money</li>
</ul>
For good measure, I’ll list a few more values:
<ul>
<li>Truth</li>
<li>Love</li>
<li>Beauty</li>
<li>Social justice</li>
<li>Happiness</li>
<li>Interpersonal relationships</li>
<li>Money</li>
<li>Power</li>
<li>Spirituality</li>
<li>Rationality</li>
<li>Imagination</li>
<li>Security</li>
<li>Excitement</li>
<li>Self-expression</li>
<li>Life</li>
</ul>
I argue that goals are always in service of values. If we look at <a href=”http://lesswrong.com/lw/sm/the_meaning_of_right/“>the terminal values in our decision tree</a>, they will be big, mysterious abstractions such as these. Why would I think this? Why not have concrete things as the terminal values of our decision tree?
Because it’s simpler. Consider this example. I can have a terminal value, “pleasure,” and then sub-goals to this terminal value: let’s say, “drinking,” “smoking,” and “eating.” On the other hand, I can have “drinking,” “smoking,” and “eating” all as terminal values, without the node “pleasure” anywhere in sight.
The latter seems rather un-parsimonious. Why have multiple specifics in my utility function, when I can have one generality that replaces them all? Scientific theories are developed by subsuming specifics in generalities. Also, I think that the case in the previous paragraph is a fairly representative example; that for every specific desirable, we can name a more general desirable, or a set of more general desirables, which it serves.
So it seems to me that the simplest theories of value are those which have mysterious abstractions as the terminal values in the decision tree. These mysterious abstractions are what I have called “values.” Our goals, in turn, are non-terminal nodes which point towards our values.
In order to achieve our values, we must figure out how to achieve them, and then do so. We must choose our goals, and carry them out. But, there is another step in the process: <i>we must figure out what we value</i>. It is this step which I wish to examine here.
I make the following claims:
<ul>
<li>Every person has a relatively stable set of values.</li>
<li>No person has a perfect understanding of what they value. There is a disconnect between what we value, and what we understand ourselves as valuing.</li>
<li>It is possible, through introspection and analysis of experience, to discover more perfectly what one values.</li>
<li>Some instances of the problem of motivation are caused by pursuing a goal which is not sufficiently in harmony with one’s values. In these cases the solution is to understand the nature of the disharmony, and change the goal, or the approach to the goal, to eliminate the disharmony. This will probably involve making new discoveries about one’s values.</li>
<li>Pursuing a goal which is strongly in harmony with one’s values does not require effort or willpower; it comes naturally.</li>
</ul>
Unfortunately, I do not have any scientific evidence to support these claims. Also, the usual qualifiers to self-help advice apply. I will conclude the post by giving several examples of how these principles have been played out in my life.
I spend a lot of time on creative projects. For years I struggled with the problem of not being able to finish anything I started. I now finish most of the things I start. This is probably due to many factors. One factor, I believe, is that I now put more careful attention into making sure that the message I have for each creative effort strongly embodies what I value. I concentrate more of what I value into each piece, and this is possible because I have a better understanding than I used to of what I value.
Most of this has to do with the detailed content of the pieces. It also has to do with the media I use. I used to write music, but I no longer do so. Now I write a lot of philosophy. This is better for me because the greatest value which music embodies is beauty, and the greatest value which philosophy embodies is truth. I value both truth and beauty, but I value truth more than I value beauty, and so philosophy is more gratifying for me.
The biggest mistake of my life so far was my attitude in a relationship with a certain girl. I spent a lot of time trying to “fix” her or “improve” her, mostly by giving her advice or knowledge about herself and her personality. The information that I gave her about herself, while true, hurt her self-esteem, and this weakened our relationship, ultimately contributing to our breaking up.
What I should have done, as I realized in retrospect, was to love and accept her for who she was, rather than trying to fix or improve her. This experience taught me the value of love. Love is something I value more than I ever knew. It has proven difficult to propagate the belief “I value love” throughout my mind, so that I often forget it. Therefore I suspect that I value love even more than I know.
(This might seem like a paradox, but it is not. When I say that “I value love more than I know,” I mean simply that I don’t feel, continuously and at a gut level, the value that I believe that love really has. This is similar to <a href=”http://lesswrong.com/lw/1l/the_mystery_of_the_haunted_rationalist/“>the rationalist who is afraid of ghosts</a>. He knows intellectually that ghosts don’t exist, but he hasn’t managed to convince the rest of himself of this.)
I practice mysticism. I use meditation, and in the past have used physical exercises and rituals, to induce <a href=”http://en.wikipedia.org/wiki/Mystical_experience“>mystical experiences</a>.
I value mystical experiences because they efficiently serve a significant number of my values. They involve intense positive emotions and aesthetic ecstasies, serving my values of happiness and beauty. They often give me new ideas, serving my value of truth, and indirectly serving my value of self-expression by providing me with things to write about. (I came up with the idea for this post while meditating this morning.)
For a long time I did not have a clear understanding of what I was trying to get out of mysticism. At times, therefore, it became a <a href=”http://lesswrong.com/lw/le/lost_purposes/“>lost purpose</a>, which did not get me anywhere.
I now practice mysticism with a clearer idea of what I am trying to accomplish. This has made my practices more effective. Mysticism involves subtle manipulations of one’s mental processes. Rather than engaging in mental processes which are simply “what you’re supposed to do in mysticism,” I now engage in mental processes which I know will lead to the results I want. This has been made possible by connecting the processes with the values they serve.
It is not only true that my mysticism has become more effective for me by connecting it with an understanding of my values; it is also true that it has become easier. Previously, I would exercise a significant amount of willpower to engage in mystical practices for an hour or two a day. Now, I meditate for three hours a day without any exercise of willpower. It doesn’t require willpower because I simply <i>want</i> to do it. I want to do it because it serves my values.
I wish to analyze in detail a particular way in which I previously went wrong with mysticism. I used to meditate in a half-lotus position. Because I was born with low hip flexibility, this position put a lot of stress on my body. I was initially in a lot of pain from this position, but I continued using it. Eventually the pain went away. Further down the line, I gave myself a knee injury from meditating in this position, which probably would have required surgery if I had continued to do this.
There are a few things that I learned from this incident. I was meditating in half-lotus to conform to my image of “spiritual purity,” based on cultural memes which I had absorbed. My knee injury made me realize that I had been on the wrong path in this respect, and I saw that I was wrong in valuing “spiritual purity.” It was not part of my value system, but something that had been imposed from outside.
My knee injury also made me realize that I had been wrong in putting my body through all of that unnecessary stress in the service of mysticism. This was one of the data points which allowed me to discover the value of self-love: of being nice to oneself.
Another data point which allowed me to discover the value of self-love came in editing the book I am working on. Editing this book has been quite painful, because I am so critical of it. Reading it makes me disgusted by it, and then I am disgusted with myself.
For example, about a week ago I came across two paragraphs which contained some statements that I felt to be wrong. I was so disgusted by these wrong statements that I could not bear to look at them. I had to spend days psyching myself up to the point that I could look at these paragraphs for long enough to delete them.
The basic problem here is that I am hyper-critical of my writing, and in criticizing my writing I am criticizing myself. This makes the editing process a terrible stressor on my self-esteem. There is a clear message about my values: I value self-love, but do not love myself enough. If I was nicer to myself, then I could edit painlessly, and I desire to edit painlessly.
A final example will illustrate once more the connection between values and the problem of motivation. I am a psychology major and a philosophy minor. I got the psychology major at the beginning of college, and the philosophy minor two and a half years later. I enjoy my philosophy classes, and find my psychology classes tedious. It is therefore easier to be motivated to do well in my philosophy classes than in my psychology classes. I chose philosophy at a time when I better understood my values than I did at the time when I chose psychology. If I had this knowledge of my values when I started college, I could have chosen a major that I would enjoy more and do better in.
I hope that the reader has found these examples illustrative. I have attempted to throw some light on the process of discovering one’s values and revising one’s goals to conform to these values. It seems to me that this is a basic tool of instrumental rationality. One cannot achieve one’s values without knowing one’s values. Furthermore, one cannot sustainably pursue anything other than one’s values, and many cases of the problem of motivation can be explained by this fact.