Was reading a post on Psychology Today by Jonathan Wai and was displeased to note the stereotypic view coming from a physics major, alias ‘Order’. While I shared my opinion on his comment, I decided I would like to share a simplistic view on this. Anyone more interested in the dividing of the sciences should take a course/read up on philosophy of science -very interesting indeed!
Why does the hierarchy exist at all? Well, the first reason would be chronological order of fields introduced in Western universities (an arbitrary reason). The second, comes from the philosophical viewpoint of Positivism -something that has been central to research within (mainly) Physics, Chemistry and Biology. When psychology started establishing itself as a solitary field, (around the 1880s/90s with Wundt/James/Hall establishing psychology laboratories) the view of a “researcher” as a general concept, was one of high dignification, as well as the rigor and high regard for positivistic science. Already here psychology decided to borrow from existing sciences to increase its credibility. Psychology copied its philosophical underpinnings (positivism), physical look (for example, introducing white lab coats for psychologists/experimenters to mimic MDs), calling participants subjects (because they were being ‘subject’ed to a process, as well as, creating a power-relation between researchER and researchEE) and other (what would seem like) details in procedure and analysis. This, I believe was a misstep like no other, because, in my opinion psychology does not strictly live up to all the positivist ideals. I can go into this in excruciating detail, but will refrain unless asked [Or see my post ‘Nonsense.’ for a short intro].
I do not believe you can force a numerical structure on something that we can barely define yet. How is ’emotion’ defined? Differently in nearly every research report about it (yes, yes, there are two general ideas about the concept, but I’m keeping this simple). The point is that it does not have a generally agreed upon experimental definition. Then how are we supposed to measure quantitatively its properties? Here, hard sciences usually lobby the; “ha! Told you they weren’t a real science, they don’t even know what they are studying!”. And this is where I claim psychology is misunderstood. What is science? Exploring the physical world, gaining an understanding of how it works. So just because psychology hasn’t yet defined all the properties within its field, it isn’t a science? Isn’t the argument the opposite? _Because_ we are, probably more so than most fields, trying to get to definitions, experimenting on flawed existing ones to make them better, and so on, should Psychology not be seen as a frontier for science?
Method. We have a hard time controlling for extraneous variables, indeed we do. Randomisation is a good help but it doesn’t cover every single bias in every single experiment published. This is in my opinion, one of the most important flaws to be aware of in psychological science. Not too seldomly, the most pressing biases for the given experiment isn’t controlled for, isn’t discussed or in any other way documented. This gives rise to the retrospective, I-got-a-significant-result-with-a-decent-effect-but-was-it-really-the-intervention-that-caused-it, question.. (..yes that was a terribly worded and awkward sentence). It is right of everyone, from any field, to critici{s/z}e psychological experiments on the grounds of bias or uncontrolled for variables -because it leads to better experiments, better defined concepts, more rigorous methods and more reliable results. But. It is wrong to dismiss psychological experimenting like a naughty child _because we can’t do better_! Yes we can! (Bob the Builder, retrieved 10:57, 18th of June 2012)
An example, by the way, of why I call these opinions simplistic. Just within psychology, we have a broad range of subfields, for example neuropsychology, and it is able to and does follow positivistic philosophy more closely than for example social psychology. It more closely resembles biology and chemistry, and often rely on those fields for answering parts of the questions posed. In either case, there are more arguments, more detailed arguments and the issue is far more complex than what the above may make it seem like.
A thought-provoking article as always Patric. I'm going to agree with your thesis broadly defined, that psychology is a science. I'm going to push back a little on how you achieve this claim.
In my opinion the greatest philosopher of science of the last 100 years, Thomas Kuhn insisted that "good science" and new theory must satisfy five basic tenets. It must be fruitful: novel predictions must be made. It must be simple: parsimony is preferable to complexity but must also be sufficiently detailed to fully explain a phenomenon. It must have broad scope: explain a broad range of abstract phenomena rather than a very narrow and rigid range of phenomena. It must be consistent: both internally and externally consistent. Finally, it must be accurate: empirically adequate but not vague. I will focus on the latter two as these are the most popular tenets of discussion and are popularly called reliability and validity in statistics.
I think you are wrong to so quickly yield that psychology is unable to measure complex phenomena such as emotion. Just because you measure something in different ways in no way means that you cannot measure it at all, but rather that you can measure it exceedingly well. The evidence is quite strong that emotion ranges along two continuums: valence and arousal. We can measure these with the PANAS for valence and the feeling thermometer for arousal. We can measure these with an fMRI for both valence and arousal or galvanic skin response for arousal. We can measure these with observations of social interactions in controlled settings such as interacting with a confederate. These diverse measures can be compared to each other for convergent validity, to other measures for divergent validity, to themselves in various permutations across or within groups for reliability and so on. These gives us a way to address the issue of reliability and validity with empirically falsifiable methods. The same can be said for such biological phenomena as the binding affinity between a protein and a specific RNA sequence. Without going into detail, there are many assays with which to measure this phenomenon that give us greater confidence that we understand the binding affinity between a protein and an RNA sequence.
My point then is that we are able to satisfy Kuhn's criterion for "good science" just as well as the "hard" sciences. You are right to argue that psychology is a frontier of science with vast potential for gaining understanding. I believe that our models of human behavior will continue to improve, becoming more reliable and valid… but also explaining broader ranges of phenomena from the brain to behavior as fields integrate and synthesize information.
Lastly, if interested I would suggest a number of articles that address well your first point, being that there are problems in psychology just as there are in every other field that we should work to address. See Gigerenzer (2009), Dawes (1994), Cacioppo (2004), Kruglanski (2001), Platt (1964), and Gergen (1973).
Interesting perspective, I too agree with the general arguments you make in the post. However..
..emotion was used as an example of a difficult to define concept. In my opinion it is not enough, and quite pointless, that we can measure 'it' unless we have some form of idea of what 'it' is. I don't mind having a working definition of something until we actually figure out what that something is, but, to say that for example emotion _is_ the measures we are currently using to get to it, then I think we are making things very easy, but misrepresentative, for ourselves (this is what I think you are arguing, correct me if I'm wrong). Physical responses have of late been associated to novelty of presented stimuli, not the valence/arousal itself (correction, arousal to a small extent, see Schupp et al., 2006). I agree to that other measures (however not restricting all) than the strictly bodily responses (galvanic skin response etc.) can, should and hopefully is, contrasted and compared for divergent and convergent validity. I would however like to keep my skepticism about correlation being treated as causation. I am not saying this is what you are doing (I think we agree on this point), but every so often you hear others argue that when picture A is shown, brain part B lights up and that means one causes the other. I say, not so fast. Firstly, it is an assumption that BOLDs (and all the other fancy abbreviated measures) represent thoughts/emotions/cognition.. it is quite the bit harder to actually prove this to be the case. I do not take the extreme stance (can't remember where I've read it atm) that we don't know what we are looking at when we see an fMRI image (or the even more so extreme stance; that we _can't_ know). I think we are getting quite far underway of understanding this -but to assume the opposite, that, brain imaging is set in stone and all clear of any errors/misunderstandings/shortcomings is equally as bad. I agree that binding affinity between a protein and a specific RNA sequence can be measured in many different ways and that this gives greater confidence, but, I claim that this is only so because we have a very specific definition of the concepts 'protein', 'binding affinity' and 'RNA sequence'..
I also agree with your conclusion however, integration and synthesisiesa..zation? =).