As an applications programmer, it isn’t the programming language that’s close to traditional science, it’s the topic they’re being applied to. Something like GUI design is pretty close to traditional science: trial-and-error, experiment under controlled (and uncontrolled) situations, theories about human behaviour, etc. The same sort of thing is true for factory control, Internet simulations, etc.
Pseudo-science has another far more negative connotation for me – it’s reporters, newscasters, and show-off scientists making dramatic fast-and-loose statements which not only present discoveries in the shallowest form, but worse, ignore other theories and discoveries in the field – causing the story to lose much of its context.
In this category put: all TV “news hour” stories, nearly all newpaper articles, most PBS “specials”, a good part of what appears in “Popular Science” and even, sometimes, “Scientific American”. In the research center I worked in “Nature” was taken to be largely reputable, but “Scientific American” was known for being a stage for big-name scientists tooting their horn (always looking for that next grant or talk show appearance!)
Pseudo-science to me is entertainment for people who mostly don’t care about theory but want to hear the “Gee Whiz, isn’t science wonderful!” stories. It’s entertainment, not science.
Sorry to hijack the thread but your email address is not available. I have a degree in Computer Science and although I would not today describe myself as a Computer Scientist I am now damn curious about your 4 lines of code. Can you post a new thread or provide a cite or something? What year did it first appear in texts?
BTW I have always said that any discipline that has to put the the word “science” in its name is not really science.
My attempt was to look at the ad hoc-ness of a hypothesis. Ad hoc modifications is changing a hypothesis after the evidece comes in so as to avoid changing the core idea.
Take a hypothesis like “Evil demons are responsible for automotive trouble.” A skeptic might check this hypothesis by looking in and around a car for evil demons. Even if none are found, the hypothesis can usually still be saved with an ad hoc motification. After none are found, a proponent might make the ad hoc modification that demons only come out at night.
A skeptic might check at night, again finding no demons. The proponent then claims that demons don’t come out every night, just occasionaly. And this can continue. Every new piece of evidence is worked into the demon hypothesis simply by making a change here and there.
There is a rational for saying ad hoc modifications are problematic. For any given data set, there are many different theories which retrodict all the data, but are mutually exclusive. This is sometimes called curve fitting: through any set of points, many different lines can be drawn. They all pass through the known points, but radically disagree on the unknown.
Thus, to circumvent this problem, we say that a hypothesis must also predict data unknown at the time of creation of the hypothesis. While many lines may pass through old data, few will be able to correctly predict new data. Presumably only the true hypothesis will be always able to predict all new data.
Having said that, ad hoc, as was pointed out, is not a perfect tool for differentiating science and pesudoscience. Science could not advance if we did not allow the formulation of new hypothesises. Sometimes even a classic ad hoc modification turns out to be correct.
As with all items I listed above, ad hoc is not a magic bullet. It is just one factor that should be considered (and some philosophers disagree here too).
Incidentally, I am happy to defend (processual) archaeology as a science. It just seems to move away from the OP.
I just want to say that this is how it is at my college. There is an Information Systems major and a Computer Science major. There are a couple classes that overlap, but they are mostly seperate entities.
The classes I’ve had for my Computer Science major are far more than learning to use Word, Excel, etc. We’ve studied algorithms and data structures of all kinds, cryptography, AI, user interface design, and numerous other topics. There’s a lot that fits under the heading of “Computer Science”.
Oh, and ftg, after teasing us like that you’d better darned well tell us what those four lines are!
I must disagree. Computer Science is more accurately described as a branch of mathematics. There is a tremendous body of pure research and theory in computer science, as with other types of math.
No, since many new discoveries lay fallow for decades before winning converts and taking over the mainstream. Examples: Krebs cycle, Chandrasekhar’s black hole concept, Goddard and space exploration, germ theory, continental drift, and many others. Were these “pseudoscience” before general acceptance, when only a few far-seeing researchers thought they were true?
Major problem: after decades of discussion, scientists themselves conclude that “science” is EXTREMELY difficult to define. Unfortunately our classes in school hide this fact. Teachers makes all sorts of noise about “The Scientific Method,” as if there is a list of procedures which, when not followed, imply that “science” isn’t happening. Wrong. When doing science, scientists don’t follow “The Scientific Method.” There is no one-line definition for science or pseudoscience. According to that “Scientific Method” list, the only true sciences are the experimental ones, so observational sciences like Paleontology, Astronomy, etc. are not science.
You could also search on keywords “nature of science”
On the other hand, it’s not too satisfying to limit ourselves to declaring that science is simply “that which scientists do.”
Here’s one solution: examine the DIFFERENCE between pseudoscience and genuine science. By this procedure we can draw a line between them and try to define each by referring to it’s opposite.
SCIENCE: truth-seeking, where even tiny bits of distortion and dishonsty are abhorrant.
PSEUDOSCIENCE: promoting an agenda, so distortion and dishonesty are perfectly OK as long as you don’t get caught.
SCIENCE: exploring the unknown. Start with a blank slate, then try to fill it in with newly discovered knowledge.
PSEUDOSCIENCE: stating a belief, then defending it against attackers. Start with confident assumptions, then search for ways to shore them up and to defeat opponents.
SCIENCE: invites criticism, and promotes fierce self-criticism.
PSEUDOSCIENCE: defensiveness, smoke and mirrors. Keep your opponent confused so he can’t give effective criticism. Never question your own assumptions, that would just give ammo to the enemy. Never even EXAMINE your own beliefs too closely, you might see faults, hurt your confidence, and lose the battle.
SCIENCE: making observations, then coming up with theories to knit them together into sensible explanations.
PSEUDOSCIENCE: creatiing theories, then defending any observations which support their predictions, while attacking any observations which contradict them.
SCIENCE: uses verbal shortcuts for quick communication: jargon heavy. Strives for extremely clear communication between colleagues who understand even tiny shades of meaning in the jargon.
PSEUDOSCIENCE: creates a false facade of “science” by intentionally coining weird terms to replace familiar words. Spread obscurity and to keep outsiders guessing.
SCIENCE: the entire goal is to understand reality
PSEUDOSCIENCE: we understand reality already. The entire goal is to convince others, or to defend our beliefs against others who wish to make us question ourselves.
A book–I don’t know the last time it was reprinted, but I believe it’s a standard–is Fads and Fallacies in the Name of Science by Martin Gardner. I have my mother’s old second (1957) edition, but I don’t think it’s an original 1957 edition. It covers the basic idea of pseudoscience and about 20 types of the more famous pseudoscience ideas. The author also suggests Daniel W. Hering’s Foibles and Fallacies of Science from 1924 and something edited by Joseph Jastrow in 1936, The Story of Human Error . I’m sure there are other books, both in general and in specifics, but it couldn’t hurt to see if you could find these three in the library, especially the Gardner.
As an environmental biologist, I must say that science doesn’t always have to take the form of hypothesis testing. Sometimes information gathering has to come before hypothesis generation or testing, and sometimes it is enough by itself.
You want to know what kinds of birds or fish or crabs use a certain habitat for example. Or you don’t know where a species of fish goes in the winter. Or you need to know the range of chemical consituents of Missouri River water. You go out and count or follow them around with radio tags, or collect the water and measure the chemicals. Sometimes you don’t have enough information to form many hypotheses before the test, and there are few statistical tests that apply. It is still reproduceable research and it does add to our understanding of the world. It is still science.
Then sometimes after you’ve gathered this data and you see relationships in the data that allow you to form and statistically test hypotheses, a posteriori. Just because you had the data before you tested the hypothesis, does not invalidate the research. It is still valid science.
Wow, this thread is getting complicated. Let me get at the easy stuff first.
I brought up how some the research of my branch involves actual proofs to contrast it as much as possible from “plumbing and carpentry”. CS is very a broad field some of which do proof like stuff and many don’t. Like I said, some are Engineers, some are nearly Psychologists (e.g., people doing Human Factors/Computer Interface stuff), etc. One guy I knew got a fellowship to learn more about Neurology and spent a year “fractionating rat brains” (shades of Walter Rosenberg/Bob Balaban).
Let me give just one example of experimental science in CS: A lot of money is being spent on figuring out how to extract information from the Web. Search engines, obviously, but also a lot of companies that see the Web as a valuable pool of raw data just begging to be turned into sellable information. “Data mining” is one of the terms to Google on if you want more. Some people think AI might be helpful here. So you have to collect a lot of web pages, follow links, try to figure out how to extract information, come up with measures that define how useful something is, run some tests with users, find out if they found your service useful, go back, get more data, refine more measures, etc. If that’s not experimental science, I don’t know what is.
Note that CS is therefore NOT a Mathematical Science. Slicing up rat brains isn’t Math. Even the Theoretical branch is NOT a Mathematical Science. Note that we use Math all the time. That’s a hallmark of all Sciences: using the Math appropriate to your field. Some people list the “P=NP?” problem as one of the top ten open problems in Math. These people are either loonies or are trying to denigrate CS. It is in no way a Math problem. It might be proven using no Math beyond the grade school level.
Short Here’s a little help for you to understand what the problem is. P and NP are two classes of problems to solve. P are those problems that have “easy” (polynomial time) solutions. E.g., sorting, matrix multiplication, shortest paths in networks, etc. NP are those problems that have solutions that are “easy” (poly. time) to check. E.g., can you put all these packages on these boxcars and have them all fit? Given an assignment of which package to which car and where, you can easily check if they fit. Finding such a solution appears to be difficult. Note that the class P is contained within NP. If it turns out that the two classes are equal, a whole lot of hard, highly sought after soutions to interesting problems become “easy”. Good news. It also would mean that all Public Key cryptosystems would be broken. Bad news. The fate of the universe hinges on the answer! (Okay, okay…)
Some points about what some people had for “CS” in high school or early college.
Learning applications like Word Processing, etc. are not in any way, shape, or form Science of any type let alone CS. No more than learning to drive a car should be worthy of credit hours. Useful does not mean public schools and colleges should give you credit hours for it.
Programming occupies the same place in CS as Arithmetic does in Math. You generally need to know the first to be able to study the second. Presumably there are Math researchers who know virtually no arithmetic, so there are probably CS profs who don’t know much programming. But generally you gotta have the first to be good at the second.
Programming used to be taught more in high school than it is now. We CS profs thought that one day we could drop programming entirely from the college curr. After all, do you get college credit for learning how to add? But schools have backslid in recent years and teach “Apps” more and programming less. (For shame!) Ideally, all students would learn how to program in elementary school, just like Arithmetic, so by High School you could start teaching them true CS. And then, we’d have avoided half of this thread since people would be a lot more knowledgable about what CS is (and isn’t).
As to my “four lines of code”. Actually I came up with an “algorithm”. Since the distinction between “program” and “algorithm” is lost on most people, I didn’t bring up the term. What appears in OS books is an implementation of my algorithm. If I wanted to identify myself further, I would have given suitable cites. But I don’t so I won’t.
Sure. There are definitely parts of CS that are experimental. But that doesn’t mean that the field as a whole is. I’m undecided as to whether I consider CS an experimental science; I need to think about it some more. There certainly are branches which are not experimental, such as algorithm analysis. For those who aren’t well-versed in CS: no one will be satisfied that your algorithm is linear time if it runs in linear time on all the inputs you could come up with, cause there’s always some that you didn’t test. Proof is necessary here. Which brings me to my next couple points…
This is a false dichotomy. A given discipline may be broad enough to incorporate experimental and theoretical aspects. Look at physics. People who do cutting edge physics research are oftentimes doing cutting edge mathematical research. What they choose to call themselves does not impact the nature of their work.
**
And why does it besmirch your discipline to be associated with math?
A problem that uses no math beyond the grade school level still may be a mathematical problem. True, it’s not highly theoretical, but I’m of the school of the thought that any problem requiring mathematical methods is a mathematical problem.
P vs. NP is a poor example, though, because at heart, it’s a question of set equality and Turing computations. Set theory is most certainly in the domain of mathematics, and Turing was definitely a mathematician (did you know he independently discovered the central limit theorem?).
I think I can see where you’re coming from, though. Theoretical computer science has a very small overlap with other mathematical disciplines (although it does exist). It tends to use ideas from other disciplines and not contribute to them, although this may only be because most people see an artificial divide between math and theoretical CS. Additionally, theoretical CS is often developed in response to some real-world problem, which seems different from higher math. But that’s not the case–most of higher mathematics was developed in response to a real-world problem, or a problem with math so far.
This seems like an issue where there’s room for reasonable people to disagree, so we may just have to do that. GD would probably be a better place to take it from this point.
Science: testing a drug to see if it works or not.
Pseudoscience: don’t you dare question whether the psychic fields around certain mineral crystals can actually promote healing. Of course they do, since ascended masters communicating through psychic channellers said so. Only a negative-minded Doubter asks for materialistic proof. You’d better not ask that healing crystals be tested for effectiveness, or you reveal yourself to be a Doubter.
Science: testing water to see if it contains traces of toxins
Pseudoscience: no doubting Scientist can tell if our expensive Vortex pendants will protect you from harmful geopathic energies. Trust us, they work great. If you don’t believe us, well here is a long list of people who swear by our product. (After all, that many honest and well-meaning people are never wrong.)
