A discussion of metagame concepts in game design

[pemerton]

the French and German academics who established the modern practice of history in the 19th century saw themselves as “scientists

Agreed, but while they developed a practice of careful observation, I think the measurement and systematisation that are characteristic of science are missing. History tends to have a particularastic element that is at odds with science.

That's not to deny that it is knowledge. Not all knowledge is scientific knowledge.

Prediction: History-as-science – if achievable – will concern itself (initially, at least) with large, long-term processes. It will involve predicting the interactions of fields with varying degrees of uncertainty. Perhaps it will resemble quantum theory more than history-as-we-understand-it-today.

Framing this in the language I have been using, this looks like a prediction about how history might try and invoke measurement.

I have seen this attempted for economic history (eg the various "wave" theories) but I don't know enough about that field to know how seriously to take those attempts.

 

[Shasarak]

7. Prediction: History-as-science – if achievable – will concern itself (initially, at least) with large, long-term processes. It will involve predicting the interactions of fields with varying degrees of uncertainty. Perhaps it will resemble quantum theory more than history-as-we-understand-it-today.

Personally I prefer History as a story, some times using a bit of science helps to get you a better story.

 

[Maxperson]

FWIW, here are my 4a.m. insomniac thoughts. I won't make any ambitious prophecies about eternity, though :

1. History is an intrinsically trans-disciplinary discipline; I would argue that in order to practice “good” history, reasonable fluency is required in sociology, psychology, anthropology, philology, climatology etc. Further, history makes routine use of data gathered by scientific methods: carbon dating, materials science, dendrochronolgy, core samples, DNA evidence, molecular archaeology etc. This does not make history science, as historical judgements still proceed from inference; what it does do, however, is provide lots of data, so hold that thought for a second.

2. Caveat: We are pattern-seeking apes. Where no pattern exists, we try to invent them.

3. There have been a number of appeals to science by historians over the past two centuries. Comte had a positivist model; the French and German academics who established the modern practice of history in the 19th century saw themselves as “scientists;” Ranke’s (completely debunked) historiographical theories; Bloch; the Annales school of historigraphy; more recently, human cycles theory and cliodynamics have attempted to model “big picture” historical processes. None of this makes history science, either; but it starts to move things in the right direction. Most importantly, it describes the recurring human desire to construe history in objective, measurable terms which can then be used predictively. Humans are also nothing, if not tenacious apes.

4. Verifying historical data is inherently problematic. Because we cannot observe history directly, all historical pronouncements are probabilistic inferences. Bayes’ theorem – and probably others, of which I have no understanding – offer ways to frame these pronouncements.

5. Further caveat: Bayes’ can also be used to justify all kinds of whacko pseudoscience. Garbage in, garbage out, and all that.

6. We need big computers to crunch lots of data.

7. Prediction: History-as-science – if achievable – will concern itself (initially, at least) with large, long-term processes. It will involve predicting the interactions of fields with varying degrees of uncertainty. Perhaps it will resemble quantum theory more than history-as-we-understand-it-today.

http://blog.yalebooks.com/2011/08/10/history-as-art-not-science/

https://vridar.org/2010/10/14/history-as-science-not-only-art-history-for-dummies-2/

 

[Ovinomancer]

Well, for what it's worth, I don't think this is just about manners.

There's a small matter of usage - if everyone in his day described Joseph Banks as a scientist, and made him President of their most important scientific society for more than 40 years, it seems odd to deny that he is one.

But there's also the issue of accurately describing a human practice. Science is a human practice aimed at generating a body of knowledge that is systematised and (in part because of that) disseminable and usable. Hypothesis formation, and testing by way of experiment (= controlled observation and measurement), is one way of generating such knowledge. Careful observation and measurement of natural phenomena is another. Such observation and measurement does at least three things:

(i) in itself, it may produce systematised, disseminable and usable knowledge (think of how important scientific cartography, surveying, etc is to much of contemporary life, from road maps and GPS to urban planning to international trade);

(ii) it may help with hypothesis formation (eg it seems unlikely that anyone would start thinking about plate tectonics without first having the data provided by scientific cartography);

(iii) it may help with hypothesis confirmation or enrichment (eg the way that Joseph Banks' collection and classification of botanical data helps show the utility of, and further develop techniques of, biological categorisation, which are themselves a, perhaps the, major part of biological knowledge before the invention of modern biochemistry over the past 50 to 100 years).​

The idea that science is equivalent to, in some confined, sense, the scientific method as that is taught in high school or first year lectures, is inaccurate as a matter of history, is misleading about the nature and richness of the bodies of scientific knowledge that have been developed over the past 400-odd years (much longer for astronomy, of course), and leads to a type of methodological fetishism that generates distorted descriptions (eg astronomical observatin gets described as "experimentation" when it obviously is not) and prioritises a certain privileged set of means (the classic chemistry or physics lab) over the actual ends of science (a body of systematised, disseminable and usable knowledge).

Again, if you say that systemic collection of observations of natural phenomenon is science, you run into the problem of baseball scorekeepers being scientists. This is why you previously added a profundity layer to your argument, which doesn't work because nothing requires any kind of science to be profound -- look at the nature of middle school science fair projects, many of whom actually are science but certainly aren't profound. Verification of previous knowledge isn't profound, but it can still be science.

Overall, your argument is outcome based: was the observation recording useful and/or did people call you a scientist. This puts the definition of science as subjective according to who's calling who a scientist or subjective as to which form of formal data collection rises high enough to be science: bird watching notes in the 1700 from far away places is science, but not last nights detailed observations of the baseball game. This cannot be right -- science cannot be defined subjectively and still be a useful tool to find truth. If you wish science to be a useful tool to find truth, then it must be process-oriented not outcome based. This should be self-evident.

Again, systemic data collection is necessary for science, it is not sufficient to be science.

You know, it just occurred to me that others may be taking my argument to be pejorative about what is and isn't science. If so, you're wrong. I've already held out that science isn't the other way to find 'truth' of our existence in this universe. To me, science is just another tool in the box, to be brought out when it applies. So saying something isn't science doesn't, for me, imply that it isn't valuable, but rather it's no subject to scientific method and subsequent verification via that method. Statistics, on the other hand....

 

[Ovinomancer]

FWIW, here are my 4a.m. insomniac thoughts. I won't make any ambitious prophecies about eternity, though :

1. History is an intrinsically trans-disciplinary discipline; I would argue that in order to practice “good” history, reasonable fluency is required in sociology, psychology, anthropology, philology, climatology etc. Further, history makes routine use of data gathered by scientific methods: carbon dating, materials science, dendrochronolgy, core samples, DNA evidence, molecular archaeology etc. This does not make history science, as historical judgements still proceed from inference; what it does do, however, is provide lots of data, so hold that thought for a second.

2. Caveat: We are pattern-seeking apes. Where no pattern exists, we try to invent them.

3. There have been a number of appeals to science by historians over the past two centuries. Comte had a positivist model; the French and German academics who established the modern practice of history in the 19th century saw themselves as “scientists;” Ranke’s (completely debunked) historiographical theories; Bloch; the Annales school of historigraphy; more recently, human cycles theory and cliodynamics have attempted to model “big picture” historical processes. None of this makes history science, either; but it starts to move things in the right direction. Most importantly, it describes the recurring human desire to construe history in objective, measurable terms which can then be used predictively. Humans are also nothing, if not tenacious apes.

4. Verifying historical data is inherently problematic. Because we cannot observe history directly, all historical pronouncements are probabilistic inferences. Bayes’ theorem – and probably others, of which I have no understanding – offer ways to frame these pronouncements.

5. Further caveat: Bayes’ can also be used to justify all kinds of whacko pseudoscience. Garbage in, garbage out, and all that.

6. We need big computers to crunch lots of data.

7. Prediction: History-as-science – if achievable – will concern itself (initially, at least) with large, long-term processes. It will involve predicting the interactions of fields with varying degrees of uncertainty. Perhaps it will resemble quantum theory more than history-as-we-understand-it-today.

I appreciate this post, it's well laid out. I disagree, of course. History doesn't need to be science to be useful. A historical model with some predictive skill isn't science, it's statistics. That's not to say that some history might not one day feed a scientific theory and be tested via the method, but I don't see how at this point without a lot of science fiction involved.

On Bayes -- excellent points. It's still statistics and can be abused just as easily. However, those abuses are far more apparent using Bayes' methods and don't hide in the choice of parameterization as well as when using frequentist methods. It's just slightly more honest about how it does what it does because you can't hide your assumptions of priors.

 

[pemerton]

Again, if you say that systemic collection of observations of natural phenomenon is science, you run into the problem of baseball scorekeepers being scientists. This is why you previously added a profundity layer to your argument, which doesn't work because nothing requires any kind of science to be profound -- look at the nature of middle school science fair projects, many of whom actually are science but certainly aren't profound. Verification of previous knowledge isn't profound, but it can still be science.

Middle school science projects can model or deploy the scientific method, but (at least in my experience) they generally are not science - they don't contribute anything to human knowledge. They are training exercises.

Overall, your argument is outcome based: was the observation recording useful and/or did people call you a scientist. This puts the definition of science as subjective according to who's calling who a scientist or subjective as to which form of formal data collection rises high enough to be science: bird watching notes in the 1700 from far away places is science, but not last nights detailed observations of the baseball game. This cannot be right -- science cannot be defined subjectively and still be a useful tool to find truth.

I don't really know what you mean by "subjective", as it is a term you deploy quite liberally!

The question of what counts a science isn't always straightforwardly answereable. Sometimes we mightn't know until after the event. Sometimes it might be contested. That can happen.

But that it isn't always easy to tell what constitutes a systematic, disseminabl and hence usable contribution to human knowledge, grounded in careful observation and measurement, doesn't meant that making such contributions is not a useful way to find the truth about the natural world, or even some aspects of the human world.

 

[Ovinomancer]

Middle school science projects can model or deploy the scientific method, but (at least in my experience) they generally are not science - they don't contribute anything to human knowledge. They are training exercises.

I don't really know what you mean by "subjective", as it is a term you deploy quite liberally!

The question of what counts a science isn't always straightforwardly answereable. Sometimes we mightn't know until after the event. Sometimes it might be contested. That can happen.

But that it isn't always easy to tell what constitutes a systematic, disseminabl and hence usable contribution to human knowledge, grounded in careful observation and measurement, doesn't meant that making such contributions is not a useful way to find the truth about the natural world, or even some aspects of the human world.

You're claiming that it's only science if if meets an standard of usable contribution to human knowledge, but you cannot define usable contribution in any sufficient manner. This means that the standard is arbitrary and subject to the opinions of it's reviewers. I'm sure you'll enact some standard of majority opinion as the arbiter of the standard, but this flies in the face of history, where wrong opinions are widely held to be true based on careful observation and measurement but paired with a wrong theory that is later overturned by an initially highly unpopular alternate theory. Your argument has the former as science until it's not and the latter as not science until it is. That's ridiculous, arbitrary, subjective, and more akin to a label of approval rather than a method of inquiry.

Science is a method of inquiry - a means with which to inquire about the universe. Your definition places it as only valid if it returns an outcome to an arbitrary standard of useful. This then discounts all inquiries that don't result in useful knowledge. You dismissal of middle school science fair projects as not-science not because they don't follow the method but because they don't meet your definition of useful contribution make it absolutely clear that your definition is about your opinion, not about inquiry. Science cannot be about the results, it must be about the process. The results are what they are, and they may not be useful, profound, or even interesting, but they are still science. Similarly, things not using the scientific method can be useful, profound, and interesting. These things are not definitional of science -- they cannot be, or many things are science, such as baseball stats (collected in a systemic and carefully defined way, and very useful and interesting to people who follow the sport).

 

[Lanefan]

Middle school science projects can model or deploy the scientific method, but (at least in my experience) they generally are not science - they don't contribute anything to human knowledge. They are training exercises.

Yet, oddly enough, assuming the projects are done correctly they are in fact contributing to science in a trivial-yet-relevant way; in that a fundamental tenet of science is that to be "proven" an experiment must be repeatable to the same conclusion...which is exactly what these projects are doing.

But that it isn't always easy to tell what constitutes a systematic, disseminabl and hence usable contribution to human knowledge, grounded in careful observation and measurement, doesn't meant that making such contributions is not a useful way to find the truth about the natural world, or even some aspects of the human world.

The bit about "contribution to human knowledge" is a red herring, in that science can also quite legitimately repeat itself (see example re school projects above) which instead of contributing to knowledge merely shores up that which is already known.

Lan-"yet in contrast this post might contribute to human knowledge without being the least bit scientific"-efan

 

[TheCosmicKid]

Again, if you say that systemic collection of observations of natural phenomenon is science, you run into the problem of baseball scorekeepers being scientists. This is why you previously added a profundity layer to your argument, which doesn't work because nothing requires any kind of science to be profound -- look at the nature of middle school science fair projects, many of whom actually are science but certainly aren't profound. Verification of previous knowledge isn't profound, but it can still be science.

I'm sort of with you on this one. As previously discussed, I'm not so quick to rule out the possibility that baseball scoring might be science of a sort, but there are reasons to say it's not beyond the question of whether it's using the scientific method or "merely" data-gathering. (For instance, baseball scorekeeping is more of a subjective endeavor than most fans like to think it is: errors are only the most common of the judgment calls a scorer has to make.) You're absolutely right that [MENTION=42582]pemerton[/MENTION]'s profundity criterion is arbitrary, and it seems unfair to say that middle schoolers are not doing science on that ground.

The thing is, language is arbitrary. Words mean what we use them to mean, and some usages are subjective or inconsistent or otherwise wacky. There are terms we use which undoubtedly have a profundity criterion in their application - "literature", for example. So simply to point out a profundity criterion is not sufficient to establish that a definition is "wrong". And in fact, as far as my own linguistic intuition goes, it seems right to say that middle schoolers are doing science but less right to say that they are scientists. And the same goes for baseball scorekeepers, even in a hypothetical world where scoring is an objective and otherwise scientific process. So, odd as it may seem, I think a profundity criterion may be in the latter word but not the former.

 

[pemerton]

I'm sure you'll enact some standard of majority opinion as the arbiter of the standard

Why would you be sure of that? It seems like a silly idea, and independenty of that, as a proposition about me, has no evidentiary foundation in anything I've ever posted.

In any event: I'm sure that some people think that toddlers in art class are producing art. I tend to think they're learning some techniques that some people can use to produce art. That's not to say that calling it "art class" is a misnomer. English is a flexible language, and the adjective can describe the output of the class, the methods it teaches, its rationale, some conbimation of those three, probably other aspetcs too that I'm not thinking of.

Likewise, school kids in science class are (i) learning some facts that are part of a body of scientific knowledge, and (ii) learning some techniques that can be used to do science. I don't think that many of them are doing science, in the sense of adding to the boy of human scientific knowledge. My take on the terminology is as above.

What about a school kid who submits as his/her project the following: a careful record of the proportion of the angle of the sun over a certain patch of backyard at noon over a series of days, and of the temperature recorded by a thermometer stuck into that patch of dirt at 10 pm on each of those days? Without more it's a pretty marginal case, but it seems as closely connected to doing science as setting up an experiment from a set of instructions that someone else wrote already knowing what the outcome would be.