Last night, I blogged a Washington Post article, "The Google engineer who thinks the company’s AI has come to life/AI ethicists warned Google not to impersonate humans. Now one of Google’s own thinks there’s a ghost in the machine."
That raises many interesting questions, and I urge you to go back to that blog post if you want to talk about anything other than this one side issue, which is raised by a single phrase in the article:
Inside Google’s anything-goes engineering culture, [Blake] Lemoine is more of an outlier for being religious, from the South, and standing up for psychology as a respectable science.
I'd like to know more about the culture at Google, which, I presume has a big impact on the larger culture. What is this rejection of psychology as a science?
Googling, I found this piece from 2012: "Why psychology isn’t science" by Alex Berezow (L.A. Times).
Psychologist Timothy D. Wilson, a professor at the University of Virginia... casts scientists as condescending bullies.... “There has long been snobbery in the sciences, with the ‘hard’ ones (physics, chemistry, biology) considering themselves to be more legitimate than the ‘soft’ ones (psychology, sociology).”
The dismissive attitude scientists have toward psychologists isn’t rooted in snobbery; it’s rooted in intellectual frustration. It’s rooted in the failure of psychologists to acknowledge that they don’t have the same claim on secular truth that the hard sciences do. It’s rooted in the tired exasperation that scientists feel when non-scientists try to pretend they are scientists.
That’s right. Psychology isn’t science. Why can we definitively say that? Because psychology often does not meet the five basic requirements for a field to be considered scientifically rigorous: clearly defined terminology, quantifiability, highly controlled experimental conditions, reproducibility and, finally, predictability and testability.
Happiness research is a great example of why psychology isn’t science. How exactly should “happiness” be defined?... How does one measure happiness?... How can an experiment be consistently reproducible or provide any useful predictions if the basic terms are vague and unquantifiable? And when exactly has there ever been a reliable prediction made about human behavior?...
[T]o claim it is “science” is inaccurate. Actually, it’s worse than that. It’s an attempt to redefine science.... That’s why scientists dismiss psychologists. They’re rightfully defending their intellectual turf.
That L.A. Times article provoked this response, by Ashutosh Jogalekar, "Is psychology a 'real' science? Does it really matter?/Fellow Scientific American blogger Melanie Tannenbaum is flustered by allegations that psychology is not a science and I can see where she is coming from."
Berezow's definition of science is not off the mark, but it's also incomplete and too narrow. Criticism of psychology's lack of rigor is not new; people have been arguing about wishy-washy speculations in fields like evolutionary psychology and the limitations of fMRI scans for years. The problem is only compounded by any number of gee-whiz popular science books purporting to use evolutionary and other kinds of "psychology" to explain human behavior. Neither is the field's image bolstered by high-profile controversies and sloppy studies which can't be replicated. But it's hardly fair to kill the message for lack of a suitable messenger. The same criticism has also been leveled at other social sciences including economics and sociology and yet the debate in economics does not seem to be as rancorous as that in psychology.
At the heart of Berezow's argument is psychology's lack of quantifiability and dearth of accurate terminology. He points out research in fields like happiness where definitions are neither rigid nor objective and data is not quantifiable... If you apply a narrow-minded definition of science then it might indeed be hard to call psychology a science. But what matters is whether it's useful. And to me the field certainly seems to have its uses.
Okay, so that refocuses us on the original belief that supposedly prevails chez Google, that psychology is not a respectable science.
Posts
45 comments:
just how slippery IS this slope?
IF we accept psychology as a "Science", What's Next?
Do we start pretending that Sociology is too?
Don't you See where that slide would lead us?
Economics? Journalism? Next thing you know.. They'd be issuing Doctorates in EDUCATION!!!
Psychology gets stupid when it thinks it's a science. Contrapositive, also true: if psychology isn't stupid it doesn't think it's a science.
Epidemiology isn't a science either. They are subjects that use parts of the scientific method to make educated guesses.
Psychology can be a respectable science, if and only if psychologists adhere to scientific principles.
Psychology is going through a bad phase at the moment, it has become afflicted with a variety of Lysenkoism, which is what happens when scientists try to bend Nature to conform to political assumptions.
There are lots of sciences that aren't really science given much of their research can't be replicated and/or falsified. Psychology does appear to be one of them to a great extent.
Richard Feynmann, physics Nobelist and most widely known to the public from the Challenger Commission, once explained the hierarchy of the sciences:
Physics is applied mathematics.
Chemistry is applied physics.
Biology is applied chemistry.
Psychology is a bunch of witch doctors.
Take this for what it is worth, or how you see things.
Personally, I consider psychology as having the same relationship to real science as alchemy had. There are elements there, but not quite yet.
Journalism isn’t a respectable profession. They make no oath to follow standards probably because their no longer appears to be a standard for ethical behavior. They profess nothing.
Funny that you circled back on that line, "Inside Google’s anything-goes engineering culture, [Blake] Lemoine is more of an outlier for being religious, from the South, and standing up for psychology as a respectable science."
It hit me when I read it yesterday that it seemed a bit weird to me to single out being religious, being from the South (are there no Southern engineers?) as two things you don't see much of at Google. (a family member of mine is apparently a one-off). But I do get that engineers, coders, programmers, and the like don't look at 'loose' science as 'real' science. If there are no numbers, how do you quantify a thing?
However, singling out a belief in psychology as respectable science is a bizarre benchmark to note and probably shines a light on how Google employees feel about themselves vs much of the rest of the world. But I'd have to say, if ever there was a tribal mentality, it would be found at Google (and Meta, Twitter, Instagram, etc.). And among the tribalists, I think you would find a LOT of people who believe or follow various psychological theories, or just go to a therapist themselves.
Not that that matters, but it's still a bizarre list of standards to note that Google-ites don't typically hold religious beliefs, come from the South (good-bye Georgia Tech, Duke, Vanderbilt, Virginia, Va Tech, Rice, etc.), and don't look at psych as a real science. I'll bet a huge number of them talk to a therapist regularly. Probably with disdain.
For a long time, colleges and universities had departments of government. In 1897, Columbia named its department the Department of Public Law and Government. But after World War II, government money began to flow to “the sciences” and departments of government suddenly began to realize the advantages of calling themselves departments of political science. Simultaneously, there was born what came to be known as “the behavioral movement in political science,” borrowing from psychology in the hope of. Justifying the name change. It did sort of work.
"Googling, I found this piece from 2023."
Wow! Google can now even see into the future.
As a PhD in Psychology, I agree that human behavior and states are hard to analyze with quantitative methods and qualitative methods are suspect. Humans have little standardization in their behavior and inner states and too many variables. People interested in psychology are not usually gifted in the high level statistical analysis needed for such complicated, feedback systems. Same with climate science.
Not to mention the fact that the company is partly owned by............a Russian.
The Alex Berezow article was from 2012. His current web page says that he has "over a decade of experience" so I guess he was new to the job when he wrote ...
That’s right. Psychology isn’t science.
Why can we definitively say that? Because psychology often does not meet the five basic requirements for a field to be considered scientifically rigorous: clearly defined terminology, quantifiability, highly controlled experimental conditions, reproducibility and, finally, predictability and testability.
https://www.alexberezow.com/
With hard sciences, the goal is to identify a problem and fix it. With psychology, sociology, and economics, people can't even agree on whether something is a problem, and if it is, how to go about fixing it.
Wasn't the movie "2001 Space Odyssey" and the more recent "I Robot" about ghosts in the machine? Even back then it sounded like an argument for evolution - order, intelligence and purpose emerging from chaos and chance.
Honestly, I find the fact that he's a Christian and a vet more surprising.
As for the whole "respectable science" idea, I'm an economist, and I'm fine with saying that I'm not a "scientist." The past thirty years have put the lie to the idea that the "hard" sciences are much better.
psychology often does not meet the five basic requirements for a field to be considered scientifically rigorous: clearly defined terminology, quantifiability, highly controlled experimental conditions, reproducibility and, finally, predictability and testability. H
The entirety of modern physics assumes that 95% of the universe is "dark matter" and "dark energy" which cannot be detected or measured. It's simply assumed to exist because it's the only way the math works out in a way that matches reality.
Biology and epidemiology have embarrassed themselves repeatedly during the COVID pandemic, as well as the ongoing transgender crap. Before that it was the decades spent pushing a low-fat, high-carb diet on the west that has resulted in a massive obesity epidemic.
About the only branch that hasn't embarrassed itself is chemistry, and I'm sure that if I dug around a bit that I'd find ways that they fail to live up to the author's standards for scientific rigor.
This isn't to say that softer sciences don't have their issues because they very clearly do. Merely that those issues do not separate the "soft" sciences from the "hard" sciences by as much as the fanboys of the latter would like us to believe.
Are teachers "scientists"?
I have a "Master of Science" degree in education from a respected "School of Education" (as respected as such a thing can be, anyway)
Does that make me a "scientist"?
If so, does it make my pronuncimientos on any scientific question better than anyone else's?
I also have an undergrad science degree in oceanography. That should make any of my scientific opinions even more unquestionable.
Full disclosure: it's a 2 year associate degree. But "undergraduate degree" sounds more science, doncha think?
If psychologists can be "scientists" why not me?
John LGKTQ Henry
Hi, what tags or switches does the Google search engine use to return news from 2023? I have a project that could benefit ...
I did wonder how psychologists decided that there were exactly five and only five main personality traits. That's a theory or a hypothesis. It's hard to think of it as a scientific fact or law. A lot depends on the definition of "main," and it must really gall scientists to hear psychologists talk about such things as if they were scientifically proven.
But I think the larger point is that the closer psychology comes to describing the real human condition, rather than its more biological or neurochemical aspects, the further it necessarily gets from anything that can be scientifically proven. The behavior of human beings is more complicated than that of atoms or molecules or cells -- language or consciousness makes it so -- and can't be reduced to equations or formulas.
Are there only five main personality traits because their first letters form the word OCEAN or CANOE, and adding another personality trait would make it harder to find an appropriate acronym?
Just substitute “biology” for psychology and look at Google.
I think many of these articles in MSM like WaPo are (1) a response to the insatiable need for “content” (2) written by and for intellectual slobs who (a) are not scientifically trained and (b) both fear/l and envy the mystery and authority that they perceive “real science” to possess and (3) structured as false binaries: something is either Science (and entitled to unquestioning respect) or Not Science (and a target for mockery).
This ignores how reality works and particularly how we humans estimate risk and reward, update our internal probability tables, build our heuristics. Rather than a binary analysis, I would look for an analog one, full of gradations. Not pass/fail but ABCD. I would call much of Psychology a C+/B: flawed but useful. We should not expect perfect (or sometimes even good) predictions but they’re better than nothing. The point, as with “real/respectable” science, is to view the whole business as provisional and subject to challenge and change, preferably after really good experimentation and observation.
Feynman’s “Cargo Cult Science” is the classic here. But if it were widely followed, WaPo et al. would be out of business.
"Wow! Google can now even see into the future."
Thanks for the heads-up. Don't know how I got it that wrong. Was supposed to say 2012. Corrected now.
Psychology has several major branches, and these are functionally independent with unique jargon. The practitioners often don't speak with each other, and some have no respect other branches.
Truly Scientific: Biological, Experimental (e.g., test and evaluation), Cognitive (laboratory; perception, memory, learning), Developmental, and--within its narrow self-imposed rules--Behavioral. They folks designed your smartphone and web pages, and are deeply involved with marketing research. These folks invented IQ tests for pre-employment evaluation 100 years ago, and then were demonized upon finding reliable racial, gender, and other differences. Swept under a rug, dismissed as bad science when the real issues were very uncomfortable findings. Some social problems have no clear solution (i.e., everything fails, why spend money?).
Flip-a-coin Sometimes Scientific: Social and Clinical. Unfortunately, Clinical (counseling; Sigmund Freud, etc.) dominates discussions and comprises 50%+ of people interested in psychology. It's full of gurus, endless compassion, wishful thinking, grand but general and abstract theories (e.g., Carl Jung), and incompatible schools of thought. Social psychology is often dressed up left wing political policy objectives disguised as science, and the field regularly experiences "crises" when their ideas don't pan out. They routinely tackle questions with dozens of interacting variables and project that have no chance of success. They fight among themselves over questions with no clear answer. When they fail they either conduct useless statistical factor analysis to cover up, or just stop talking about it. The most reliable branch of Social Psychology is Personality, as it relies on data and meshes well with underlying cognitive factors.
The Hard Sciences are sometimes known as the "simple sciences," for Chemistry, Physics, and Biology have the luxury of looking at single factors and reaching incredibly reliable conclusions. You can either tackle the hard questions and more often fail, or ignore the hard questions and miss half of all potential science questions. The glass is half full or half empty.
Most psychologists used to be trained in scientific methods and were exposed to the philosophy of science. However, that objective training was replaced by the leftists who strongly influenced todays woke academics. Terms like hermetic theory and postmodernism infiltrated Psychology in the 80's and have really damaged the field.
Time to repost this great article on the history and fatuosness of schools of journalism.
It's Michael Lewis so even if you think he's full of shit, you know it will be interesting and well written.
https://newrepublic.com/article/72485/j-school-confidential
John LGKTQ Henry
Also for those who have not re-read it recently
https://sites.cs.ucsb.edu/~ravenben/cargocult.html
Thanks for the reminder Owen.
John LGKTQ Henry
Enigma @ 9:12: very useful anatomization* of the “science.”
*I use the word in its original sense: a complex structure dissected for ruthless examination…
Enigma @ 9:12: very useful anatomization* of the “science.”
*I use the word in its original sense: a complex structure dissected for ruthless examination…
Howard said...
Epidemiology isn't a science either. They are subjects that use parts of the scientific method to make educated guesses.
Agreed, mostly (I'm part Epi, btw). I tend to think of these things more on a sliding scale of Strong----Weak science. Epi and Psych and such are inherently weak simply because there are so few controls. One can't, for example, completely control the diets and lifestyles of thousands of people over several years just so you can nail down, say, the effects of trans fats on cancer rates. And even if you could, it's probably not generalizable to the real world where hundreds of other factors will mess things up.
Epi studies are just plain inherently weak science and should be treated as such. We saw this with the mask debate (or wanted-to-be lack thereof), a bunch of incredibly weak studies were held up as somehow definitive. They're not, although they're not useless either. To be sure, in their day to day work, most Epi's acknowledge all this; trouble is, when their politics conflicts, they're go with the politics every single time.
A bridge doesn't care how the designers felt about themselves, it either stands or falls based on the quality of the science/engineering used to build it.
Howard doesn't like epidemiology and I can understand that. However, real epidemiology (unlike the crap we have witnessed the past two years) is applied statistics. The statistics depend on the data and the data has been corrupted by politics. Similar to Climate Science" which has mastered the milking of the federal government but cannot predict the weather tomorrow. Crap, accurate to the fourth decimal place is still crap.
"Thanks for the heads-up. Don't know how I got it that wrong."
Freudian slip?
Whatever Google is socially, culturally, or politically; it is still a national treasure. The search engine/ ad business at least pays for the mind blowingly effective machine learning algorithms they create, and basically give away. The Inception and Resnet neural net structures are simple but so remarkably effective they are works of art. So many things about Google are weird to me, but I am amazed at how competent they are technically.
Tregonsee writes, "Richard Feynman, physics Nobelist and most widely known to the public from the Challenger Commission, once explained the hierarchy of the sciences..."
Feynman had at least one gripe against psychology that he took some pains to explain. A grad student working for an MS in psychology attending one of Feynman's very popular "physic for poets" lectures, told him about having his master's project experiment proposal rejected by his advisor. That's already been done, you need to test an original hypothesis, said the advisor according to the story.
Feynman was livid by that rejection of fundamental science -- the grunt work which is 90 percent of valid research.
"Economics? Journalism? Next thing you know.. They'd be issuing Doctorates in EDUCATION!!!"
That only makes sense, as the word "doctor" comes from the Latin for "teacher."
I firmly believe that the social media companies have the very best psychologists in the world working for them. I think that psychology informs all their algorithms. They know what works and what does not work. I think if you spend any significant time on a platform, they could pull your file and generate a more accurate psychological profile than even you, your closest friends, or your doctors.
I have had a Gmail account for 20 years. They know who my friends are, they know my IQ, they know my friends' IQ. They know my interests. They know what Youtube videos I watch. They know whether I drink or use drugs, and if I do, they know what I read, watch, and listen to when doing that and not doing that. They know what psychological maladies I have or are predisposed to, and which ones I do not. They know what medical conditions I have; they know which ones I am worried about having. They know all my hobbies and interests. They know what I look like, how old I am, and what I sound like. They know how long I spend on the computer every day. They literally know everything about me, like God.
They have all my comments on this blog.
All of that is just the raw data. Psychology then informs their algorithms at every point.
Facebook, Google, Twitter, Apple, Microsoft...they employ the best, most effective psychologists in the world, and they know which parts work and do not.
They have huge teams of not just pure psychologists, but also human computer interaction specialists -- user experience specialists, UX -- who do A/B testing on every page, conduct focus group research, analyze logs, iterate constantly, and increment results. Psychology is a qualifying degree for this field. It is one of the highest paid, and it is the single most respected field in all of tech.
Many of you read/listen to Scott Adams. Think of what he knows about the psychology of persuasion coupled with knowing everything about the target of persuasion and then built into a machine learning algorithm. Except, they use large teams of outstanding psychologists, not just one.
Let me use a tangential example. Futurism seems like a very iffy proposition, but Ray Kurzweil is pretty damn good at it. When did you last hear from Kurzweil? What happened to him? The point is, they hire the best people to inform their core business.
Their core business is control.
Scientists learn to predict and control the phenomenon they study. Psychology is the study of human behavior.
Don't be fooled into thinking they don't take psychology seriously.
Science cannot discern between origin and expression. The "ethical" alternative denies women and men's dignity and agency, and reduces human life to negotiable commodities.
Many interesting and informative comments. Freud's The Future of an Illusion is perhaps just as applicable to psychology as to religion. We believe in things that we have to believe in. Survival and a zestful life are only infrequently based on logic. There are any number of damaged people who have need of a higher power to stabilize the chaos or provide a ladder out of their morass....I'm sure that Aimee Semple McPherson cured more people than Sigmund Freud, but many psychotherapists have provided real help for their patients. It's not science, but it occasionally works. Back surgery helps for some patients and not for others. That's the way it goes....I just hope that this talk of sentient AI's does not inhibit the growth and development of sex robots. Sex robots seem to be the only fair way of negotiating the imbalances between our higher and lower impulses.
Is psychology a respectable science?
In order for what the reporter writes about Google to be true:
* There have to be a large number of Google employees who have an opinion on that
* The majority opinion of those who HAVE an opinion about that has to be that the answer is "clearly, no"
* And of those people, enough of them have to take it as an article of belief of sufficient importance that if you think it is respectable, you are regarded as way out of step with an important company norm
Taken together, that all seems pretty unlikely to me. But if it's true, that's a really interesting fact about Google employees.
At the heart of Berezow's argument is psychology's lack of quantifiability and dearth of accurate terminology. He points out research in fields like happiness where definitions are neither rigid nor objective and data is not quantifiable... If you apply a narrow-minded definition of science then it might indeed be hard to call psychology a science. But what matters is whether it's useful. And to me the field certainly seems to have its uses.
It sure does! Lawyers in criminal trials refer to psychologists as "whores", because "you can always find one to give you satisfaction".
Beyond that?
1: Do tell us what actually positive to society "uses" psychology haw provided
2: Tell us why "useful" and "scientific" are supposed to be synonyms
Psychology is not a science? What's next, psychiatry is not medicine?
Psychology used to think that gender dysphoria was a psychological problem, but now it's just a stepping stone to medical adventurism.
I was truly shocked a few years ago when I realized the National Science Foundation (funde by our tax dollars) gives grants in Political Science.
This is a list of grants
https://www.nsf.gov/awardsearch/simpleSearchResult?queryText=political+science&ActiveAwards=true
Up until then I had some doubts about the feasibility to reducing Federal Spending by 5% a year until the budget is balanced, but that convinced me.
I will contribute to any candidate who promises the Federal budget will be less than the year before. Unfortunately there is only Rand Paul.
I could have this wrong, but when I read that, I immediately thought of James Damore, the google engineer who wrote the memo explaining why chicks can't code, and got fired for his troubles. He quoted extensively from published psychorubbish in his memo. I suspect that "believes psychology is a respectable science" is now google-speak for "believes it is possible to say what a woman is".
Re. Dave's comment at 6/12/22, 11:29 AM:
On target!
Scientology?
Post a Comment