Chapter 7 Chapter 3 Operationalism and Essentialism - "But, Doctor, what does that really mean?"

Do physicists really understand what gravity is?I mean really.Do they know what the term "gravity" really means?What is its inner nature?What do you ultimately mean when you speak of gravity?After all, what is it? Questions like these reflect a scientific view that philosopher Popper called "essentialism."This view holds that only a good scientific theory can provide a final explanation of phenomena from the perspective of intrinsic nature or essential attributes.Those who support this view also usually believe that any theory that cannot provide a final explanation of phenomena is useless, that such a theory cannot reflect the real internal situation, and cannot reflect the essence of the way the world exists.In this chapter, we discuss why, instead of answering essentialist questions, science advances through the operational definition of concepts.

In fact, scientists do not attempt to acquire the kind of knowledge that essentialists seek.In this sense, the correct answer to the question posed at the beginning of this chapter is that scientists do not know what gravity is.Science does not attempt to answer "ultimate" questions about the universe, as Peter Medawar (1984) wrote: (However,) Even if science cannot answer the ultimate question, it does not mean that other answers must be accepted; nor can it be taken for granted that since such ultimate questions can be asked, they must be answered.As far as our present understanding is concerned, such questions are unanswerable. (p.60)

(But, in the end) the potential of science is limitless in terms of the kinds of questions it can answer...nothing can stop or halt its progress, except moral flaws such as lack of courage. (p.86) One reason why scientists question those people, theories, or conceptual systems that claim to give absolute answers to ultimate questions is that scientists believe that ultimate questions cannot be answered.Scientists do not claim that they can provide perfect knowledge; the unique advantage of science is not that it is an infallible process, but that it provides a way to eliminate errors, which can continuously eliminate errors in our understanding.Furthermore, claims and practices that claim to be perfect or absolute knowledge often hinder people's exploration.Free and open exploration of knowledge is a prerequisite for scientific activity.Scientists are always skeptical of claims that they have found the ultimate answer to a question.

The essentialist attitude often manifests itself in a preoccupation with the definition of terms or concepts before seeking knowledge. "But we must first define our terms" is a slogan often used by essentialists. The idea of ​​"what is the real meaning of a theoretical concept?" seems to imply that before a word can be used as a theoretical concept, we must have a full and clear understanding of all the underlying linguistic issues involved in its use. understanding.In fact, it's the exact opposite of how scientists work.Before studying the physical world, physicists don't bother to discuss how to use the word "energy" or whether the word "particle" really expresses the essential meaning we mean when we discuss the basic components of matter.

In science, the meaning of a concept is determined after, not before, the phenomena to which the term relates have been studied to some extent.A precise conceptual term comes from the kind of interplay of data and theory inherent in the scientific process, not from debates about language usage.Essentialists engage us in endless verbal arguments that many scientists firmly believe detach us from the substance of things.For example, to the question "What is the real meaning of the word life?" the two biologists responded "There is no real meaning, it is just a usage that is good enough for our job as biologists, not subject of debate or refutation" (Medawar & Medawar, 1983, pp.66-67).In short, the aim of the scientist is to explain phenomena, not to analyze words.The key to progress in all scientific disciplines is to abandon essentialism and embrace operationalism.This is exactly what we explore in this chapter.No other discipline illustrates this question better than psychology.

So, where does the meaning of concepts in science come from, if not from linguistic disputes?What are the criteria for the correct use of a scientific concept?To answer these questions, we must discuss operationalism.It is essential for constructing theory in the sciences, and especially for evaluating theories and ideas in psychology. Although operationalism comes in many forms, it is most effective for users of scientific information when it is thought of in the broadest light. "Operationalism" is simply the idea that concepts in a scientific theory must be grounded in, or related to, observable events that can be measured.What connects concepts to observable events is the operational definition of the concept, which makes the concept public.An operant definition separates concepts from personal feeling and intuition and allows them to be tested by anyone performing measurable operations.

For example, defining the concept "hunger" as "a bad feeling in my stomach" is not an operational definition because it is tied to a person's experience of a "bad feeling" and therefore cannot be known to other observers .Instead, a definition that involves some measurable food deprivation time or a physiological indicator like blood glucose level is operational because it includes observable measurements that anyone can perform.Likewise, psychologists disagree with defining "anxiety" as "the discomfort and tension I feel from time to time," and must instead define the concept in terms of manipulations like questionnaires and physiological measures.The above-mentioned definition is a personal interpretation of the physical condition, which cannot be replicated by others; while the latter is an interpretation of the concept in the field of public science.

In science, it is important to realize that a concept is defined by a sequence of operations rather than a single behavioral event or task.Instead, subtly different tasks and behavioral events are usually aggregated into a single concept (we'll talk more about aggregated operations in Chapter 8).For example, educational psychologists define the concept of "reading ability" in terms of performance measured using standardized instruments such as the Woodcock Reading Ability Scale (Woodcock, 1998).The total reading ability score measured by this scale incorporates a number of ability indicators measured by different subscales.These subscales measure slightly different abilities, but all relate to reading.For example, read an article, think of a suitable word to fill in the blank in the article, write a synonym for a word, spell a difficult word independently, etc.Performance on all of these tasks collectively defines the concept of "reading ability."

Operational definitions force us to think seriously and empirically about how we define a concept, empirically, in the sense that we base it on our observations of the real world.Imagine that we want to give an operational definition to a seemingly fairly simple concept of "typing ability".Imagine you did this to compare the pros and cons of two methods of teaching typing.Think about all the decisions you have to make.Of course, you want to measure "typing speed".But how long should a paragraph be?An essay of just 100 words may be too short, while an essay of 10,000 words seems too long.So how long is good?How long does typing speed best fit our theoretical construction of the concept of typing ability?What type of articles are used for testing?Will the article contain numbers, formulas, and unusual spacing?How do we handle errors?When we measure typing ability, both time and errors seem to be taken into account, but how do we calculate an overall score if we take both indicators into account?Do we want time and errors to have equal weight, or is one more important than the other?Finding a good operational definition will force you to think about all of this; it will make you think hard about how you conceptualize "typing ability".

To be useful, operational definitions of concepts must be both reliable and valid.Reliability refers to the consistency of the measurement tool.Whether you get the same measurement if you test the same concept multiple times.The scientific concept of reliability is easy to understand because it closely resembles the common-sense definition as well as the dictionary definition: "a property of any system that always produces the same result". Just imagine, how would a layman evaluate whether something is credible?Imagine a person who has to catch the bus from New Jersey to work in Manhattan every morning.According to the schedule, the bus should arrive at the station where this person is waiting at 7:20 am every day.During a week, we can say that the arrival time of the bus is plausible in that week if the bus arrives at 7:20, 7:21, 7:20, 7:19, and 7:20, If the next week the car arrives at 7:35, 7:10, 7:45, 7:55, and 7:05, then we can say that the car's arrival time during that week is highly implausible.

In the scientific domain, the reliability of an operational definition is assessed in a similar manner.If we measure the same concept multiple times and the results are approximate, then we say that the measurement tool exhibits high reliability.If on Monday, Wednesday and Friday of the same week, different versions of the same IQ test are used to measure the intelligence of the same person, and the scores obtained are 110, 109, 110 respectively, then we can say that this IQ test is very reliable .On the contrary, if the three test scores are 89, 130 and 105 respectively, then we can say that this IQ test does not show high reliability.There are specialized statistical methods for assessing the reliability of different types of measurement instruments, which are covered in all standard methodological introductory texts. But remember, reliability refers only to consistency and nothing else.For an operational definition, reliability alone is not enough. Reliability is a necessary but not sufficient condition.For a good operational definition, the operation must be proven to be a valid measure for the concept. The term "construct validity" refers to whether a measurement instrument (operating definition) measures what it is supposed to measure.Professor Paul Cozby (Cozby, 2006) told us a funny example of reliability but no validity in his methodology textbook.Suppose you want to test your intelligence. The tester asks you to stand on a testing device similar to a shoe sizer, and the device gives a reading.Of course, you'd think it was a joke.Note, however, that this measurement tool can exhibit many types of reliability that are discussed in methodology textbooks.The instrument will give fairly consistent readings on Mondays, Wednesdays, and Fridays (this is called "test-retest reliability"), and it will give the same readings regardless of who operates it (called "rater confidence"). Spend"). The problem with using a shoe sizer to measure intelligence is not reliability (it is), but validity.It's not a reasonable way to measure the concept it's supposed to measure (intelligence).One of the pieces of evidence that concludes that it is not a valid way of measuring intelligence is that we found no relationship between it and some other variables thought to be associated with intelligence.Shoe sizer measurements have nothing to do with academic achievement, with neurophysiological measures of brain function, with success in the workplace, with measures of information processing efficiency proposed by cognitive psychologists; instead, true intelligence tests are linked with all of these are all related (Deary, 2000; Geary, 2005; Lubinski, 2004).In the field of psychology, a real intelligence test must take into account both validity and reliability, while the shoe sizer test of intelligence has only reliability but no validity. At this point, you may be wondering whether other combinations of reliability and validity are possible.So let me reiterate our position.In the operational definition, we seek both reliability and validity, so a combination of high reliability and high validity is the ideal goal.We just discussed the shoe size IQ test in order to demonstrate that high reliability and low validity are useless.The third case is low reliability and low validity, which is absolutely useless and therefore not worth discussing.But you might wonder about the fourth and last possible combination: what about high validity and low reliability?The answer is that, like the low validity and high reliability example (the shoe sizer example), this combination is useless.In fact, it would be more accurate to say that such a situation could not have arisen at all.Because, if you can't make credible measurements, you simply can't claim that the measurements are valid. The link between concepts and observable operations varies widely, both directly and indirectly.Few scientific concepts are defined almost entirely by observable manipulations.Most concepts are defined in a more indirect way.For example, the use of some concepts depends both on a sequence of operations and on special relationships between it and other concepts.Finally, there are concepts that are not defined directly by observable operations, but by their relationships to other concepts.Such concepts are sometimes called "latent concepts" and are very common in psychology. As an example, many studies have focused on the so-called Type A behavioral pattern because of its association with the incidence of coronary heart disease (Austin & Deary, 2002; Curtis & O'Keefe, 2002; Matthews, 2005; Smith, 2003; Suls & Bunde, 2005).In Chapter 8, we will discuss Type A behavior patterns in more detail.However, the point to be made here is that Type A behavioral patterns are actually defined by a series of secondary concepts: a strong desire to compete, latent hostility, rush behavior, a strong drive to achieve goals, and so on.However, each of the concepts used to characterize Type A behavioral patterns also requires operational definitions in their own right.In fact, researchers have gone to great lengths to operationally define each concept.The point of our discussion is that Type A behavioral patterns are a complex concept that are not directly defined by operations.Instead, the concept is linked to a number of other concepts that each have operational definitions. Type A behavioral patterns provide an example of an indirect operational definition.Although different concepts have varying degrees of connection to observable operations, all concepts derive meaning to some extent through their connection to observable operations. The definition of a scientific concept is not fixed but changes as relevant observations are enriched.It is very important to be aware of this.If the original operational definition of a concept proves theoretically invalid, that definition is discarded and replaced by another set of defined operations.In this way, with the accumulation of relevant knowledge, scientific concepts continue to evolve and their abstraction gradually increases.For example, for a while it was thought that the electron was a tiny negatively charged sphere orbiting the nucleus of an atom.Today, electrons are regarded as probability density functions with wave-like properties under certain experimental conditions. In the field of psychology, the development of the concept of intelligence provides a similar example.At first, there was only one strict operational definition of intelligence: intelligence is something measured by tests of mental function.The concept has been enriched and refined as experimental evidence has accumulated that intelligence is associated with academic achievement, learning, brain injury, neurophysiology, and other behavioral and biological variables (Deary, 2000, 2001; Geary, 2005; Lubinski, 2004; Sternberg, 2000; Sternberg & Grigorenko, 2002; Sternberg & Kaufman, 1998; Unsworth & Engle, 2005).It now appears that when defining the concept of intelligence, it is best to use a high-level construct to define it through a variety of more specific information processing operations.Of course, these hypothetical information processing processes should have a more direct operational definition, which can be expressed by measurable indicators. Concepts in the theory of human memory develop in the same way.Modern psychologists rarely use general concepts like "memory" or "forgetting"; instead, they measure memory sub-processes that can be clearly defined, such as short-term auditory memory, symbol storage, semantic memory, and episodic memory.The traditional concepts of "remembering" or "forgetting" are refined through more explicit operational concepts. Thus, the usage of theoretical terms evolves in scientific practice rather than in debates over the meaning of words.This is the most striking difference between the operational attitude of science and the essentialist pursuit of absolute definitions.The neurologist Norman Geschwind (1985) described this distinction as follows: "I think one thing you can learn from the history of medical development is that many people think that studying a problem The best way is to first define the problem and then go to work on it. This idea is proven wrong again and again, because you find that you can't define the problem correctly until you know the answer" (p. 15). The philosopher Paul Churchland (Churchland, 1988) emphasized that in science, concepts are defined not by words but by observations and other concepts related to them: As scientific concepts evolve, concepts are often intertwined with many different theoretical systems and acquire multiple operational definitions.This happens not because there is something wrong with the concept itself.For example, many people consider psychology to be distrustful because many important theoretical concepts in psychology—such as intelligence—can be operationalized and conceptualized in more than one way (Sternberg, 2000).But this situation is not unique to psychology, nor is it a hopeless or helpless thing to do.In fact, in the field of science, this situation is ubiquitous.For example, "heat" can be conceptualized in terms of both thermodynamic theory and kinetic theory.Physics has not been vilified for this.Consider an electron, many of its properties are explained in terms of waves.However, some other properties of it are better understood if you think of it as a particle.So far, no one has proposed to abandon physics because of these multiple conceptualizations. Many people can understand the need for operationalism when thinking about physics or chemistry.They knew that if scientists were going to talk about a certain type of chemical reaction, energy, or magnetic field, there had to be a way to measure it.Unfortunately, when people talk about psychology, they often fail to recognize the need for operationalism.Why is the obvious not equally recognized: in order to be useful explanatory systems in scientific theories, psychological terms must be operationally defined, directly or indirectly? One of the reasons why people misunderstand psychology is what is called "presupposition bias" in psychology.We mentioned this problem in Chapter 1.Geology is not studied out of an attachment to some belief about the nature of rocks, but in psychology the situation is very different.Each of us has an intuitive theory of personality and human behavior that we use to "explain" our own behavior as well as that of other people.Theoretical concepts (such as intelligence, aggression, and anxiety) are embedded in all of our personal psychology theories.It is therefore natural to ask why we must accept some other definition.While this attitude is superficially reasonable, it is a formidable obstacle to any science that seeks to understand human behavior, and a source of public confusion about psychology. The biggest source of misunderstanding, and the biggest obstacle to the accurate representation of psychology in the media, is that many of the professional concepts in psychology are expressed in everyday terms.These everyday terms open the door to the spread of a great deal of misunderstanding.The layperson seldom realizes that when psychologists use the terms "intelligence," "anxiety," "aggression," "attachment," and the like as theoretical concepts, they don't necessarily mean what the public usually means . The nature of this distinction can be seen from the previous discussion of operationalism.When terms such as "intelligence" and "anxiety" are used in psychological theory, their correct use is determined by their direct or indirect operational definition.Those definitions are often highly technical, often have specific meanings, and differ in many ways from the everyday use of these terms.For example, when we hear the phrase "first principal component from factor analysis of a large sample of cognitive tasks," many people don't realize that it is an operational definition of the term "intelligence." Likewise, if a layman uses the term "depression," it means "feeling shitty."In contrast, the dedicated definition of depression in the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 1994) takes up more than 12 pages and is not the same as "feeling bad "There's a big difference.What clinical psychologists call depression is not what laymen call depression (Hollon, Thase, & Markowitz, 2002).The same problem exists in other fields of science, although perhaps not as much as in psychology.Recall the previous discussion of the concept of "life."The problem, as Midwar et al. (1983) pointed out, is that, like other technical terms in science, the word "life" comes from people's everyday language, but its meaning in scientific situations is far different from that in ordinary conversation (p. 66). Physicist Lisa Landau (Randall, 2005) has discussed how such questions hinder public understanding of physics.She points out that the term "relativity" in Einstein's theory of relativity is understood by the public as "absolutely nothing, because everything is relative", when in fact, the theory means the exact opposite!Landau pointed out that, in fact, Einstein's theory of relativity holds that: "Although the implementer of measurement needs to rely on his reference object and frame of reference when measuring, in fact, the physical phenomena he measures always have a constant property, this property beyond the observer's particular frame of reference".Einstein's theory of relativity is actually looking for constant properties of physical phenomena.In fact, Einstein also thought that it would be more appropriate for his theory to be named "constancy theory".However, the status of the word "relativity" was already too deeply rooted in people's minds at that time (p. 13). Landau goes on to point out that even in physics, "ambiguous word choice is at the root of some misunderstandings, and scientists often use colloquial terms. They assign specific meanings to these words, but people without formal training are not may conceive of this meaning” (p. 13).The same is true in psychology.Misunderstandings often arise between psychologists and laypeople when they use the same words to mean different things.Such troubles might be less disturbing if new words were coined to describe mental concepts.Sometimes there are words like "erg" and "joule" for physicists and "dissonance" and "code" for psychologists, words that aren't made up but are relatively rare in everyday language , thus preventing confusion. "But," a layman might object, "why do psychologists torture us like this? Why are there so many new terms, highly specialized definitions, unfamiliar words? Why do we need these? What if the definition of the concept is not approved by them?" Here, we look at an example of a serious misinterpretation of psychological research—a misunderstanding that often appears in media coverage of psychological research.A national newspaper reported on a 1996 meeting of the American Psychological Association under the headline "Can you rephrase it in layman's terms?" ".The paper mocked a paper presented at the conference titled "Explaining the Joint Factor Analysis of Pair and KAIT Using Gf-Gc Theory".Although the reporter stated that he "did not venture to guess what the headline really meant," almost all trained psychologists could understand that the headline was about a new development in the theory of intelligence testing.indeed so. The Gf-Gc theory is a technical advance in the theory of intelligence, and there is no reason for a reporter to have heard of the concept—just as the reporter would not be expected to know the details of an elementary particle recently discovered by physicists.Sometimes, however, journalists' ignorance of scientific terminology (which is completely understandable) has a negative impact on modern psychology. Let's look at the crux of the problem.The first step in addressing it is to emphasize a point we have already discussed: Operationalism is not unique to psychology, it is characteristic of all sciences.Most of the time, we accept it easily and understand its obvious nature.If a scientist studies radioactivity, we take it for granted that he must have a way to measure this phenomenon—that other researchers can use this method to obtain the same results.Operational definition thus leads to the openness of science, which is one of the key features of science.If two scientists agree on the same operational definition, one can use it to replicate the other's results.But what seems obvious in other contexts is not so clear when we talk about psychology.People often don't realize the need for operational definitions of concepts like "intelligence" and "anxiety" because we use these terms all the time, don't we "know" what they mean? The answer is: "Yes, we really don't know what they mean"—not in the sense that scientists have to know, but in the sense of the public.A scientist must "know" what intelligence means by defining precisely a method by which other experimenters can measure the concept in exactly the same way and come to the same conclusions about it.In its clarity and precision, this is very different from the vague language used in everyday conversation to achieve mutual understanding. The problem that plagues all intuitive (non-empirical) belief systems is the same problem that can arise from relying too much on what we "know".What you "know" about a certain thing may not be the same as what Zhang San and Li Si know. How do we decide who is right?You might say, "I feel strongly that what I know is true," but what if John's views differ from yours, but feel stronger than yours?Johnny has a different point of view from both of you and claims to be right because he feels even more strongly than Johnny. This simple little passage is merely intended to illustrate a basic feature of scientific knowledge, which has been an important humanizing force in human history: in science, the truth of knowledge does not depend on the individual when he makes a claim. degree of certainty.All belief systems based on "intuition" have a common problem, that is, they lack a mechanism to judge which is right and which is wrong when conflicting views arise.Because everyone intuitively thinks they are right, but when people's intuitions conflict, how do we decide who is right?Sadly, history shows that such conflicts often result in power struggles. Some falsely claim that the operational orientation of psychology dehumanizes people, and that we should base our ideas about humans on intuition.Psychologist Donald Broadbent argued in 1973 that a truly human view is one that bases theoretical views about humans on observable behavior rather than on the intuition of the theorist : The humanizing power of science, then, is to bring knowledge into the open, so that any conflicting views can be tested in a mutually acceptable way.Recall the concept of "repetition" mentioned in Chapter 1.This allows us to choose from among theories in a peaceful way that we all agree in advance.The public nature of science rests heavily on the idea of ​​operationalism.By defining the concept operationally, the concept enters the public domain—anyone can criticize, verify, improve or deny it. Psychological concepts should not be based on someone's personal definitions, which may be uncommon, personal, or vague.For this reason, psychology must reject all personal definitions of concepts (just as physics rejects a personal definition of energy, and meteorology rejects a personal definition of clouds), and sticks to definitions that are known to the public, This definition defines concepts in terms of operations and ensures that those operations can be carried out by anyone with the proper training and equipment.Psychology does not shut out the layman in terms of abandoning personal definitions, but opens the field to the public, as all disciplines do, in search of universal, publicly accessible, shared by all. Use of knowledge. Such publicly available knowledge can be used to solve human problems only when concepts are based on operational definitions and do not focus on the literal meanings discussed by essentialists.For example, Monk (1990) described how the concept of "traumatic shock" became problematic in medicine during World War II.Some physicians base the diagnosis on the high concentration of red blood cells, which they attribute to the infiltration of plasma from the blood into the tissues.Other physicians diagnose "traumatic shock" based on low blood pressure, pale skin, and a tachycardia.In other words, physicians have inconsistent (even personal) operational definitions of the concept, and so Dr Grant of the British Medical Research Association has suggested that the concept of "traumatic shock" should be discarded , and the term is not used in detailed observation of the wounded . . . the lack of a common basis in diagnosis prevents the effects of various therapeutic measures from being assessed" (Monk, 1990, pp.445-446). In other words In other words, this concept does more harm than good because of the lack of a universally agreed-upon definition that makes it public knowledge (that is, widely shared and recognized). Sometimes, in scientific fields, a change in the meaning of a concept can lead to a conflict between the scientific understanding of the concept and the layman's understanding.Farber and Churchland (1995) discuss a situation concerning the concept of "fire".The classic concept defines fire as follows: "It is not only the burning of carbon-containing substances, but also includes the activities on the sun and various stars (actually nuclear fusion), lightning (actually electric white-heating phenomenon), arctic Light (actually spectral emission), the flash of a firefly (actually phosphorescence). In the modern conceptual system, these phenomena have nothing to do with oxidation, so they do not belong to the same category as wood burning. In addition, some phenomena were originally thought to be related to Processes that had nothing to do with combustion (since the exotherm was considered an essential feature of combustion at the time)—such as rust, corrosion, and metabolism—turned out to be oxidation phenomena” (p. 1296).In conclusion, the principle of oxidation is what connects campfires to rust, and distinguishes lightning from them.For scientists, this may be a sign of progress, but it leaves laymen confused and at a loss. Another reason many people abandon the operationalist perspective when they approach psychology is that they want to find essentialist answers to these questions.Whether this is done because psychology has recently separated from philosophy, or because the public knows less about psychology than about other disciplines, is unknown.In a sense, however, it doesn't matter.The end result is the same.Psychology is expected to provide ultimate answers to these complex questions that no other discipline can answer. Think back to the question posed at the beginning of this chapter: What does the term "gravity" really mean?What is its inner nature?What exactly do we mean when we speak of the term gravity?Most people think that these questions require absolute knowledge, an understanding of the underlying nature of phenomena, and that current theories of physics cannot provide answers to such questions.Anyone familiar with popular reading on the development of the physical sciences over the last few centuries is aware that gravity is a highly complex theoretical construct, and that its conceptual and operational relationships are in constant flux. However, if all the "gravity" in the above questions are replaced with "intelligence", a miracle will appear.现在那些问题立刻被赋予了重大意义。它们看起来是那么自然和富有深意，它们就是在寻求一个终极答案。可是当心理学家给出和物理学家一样的答案，即“智力是一个复杂的概念，它的意义是由测量它的操作以及它与其他概念之间的理论关系来界定的”时，却会被鄙视和指责为回避真实问题。 心理学所面临的一个难题就是，公众要求心理学去回答本质主义问题，而通常其他科学家并不需要回答类似的问题。这类要求常常导致人们贬低心理学领域已经取得的进步。尽管这类要求不能阻止这一领域自身的发展——因为就像其他科学家一样，心理学家无视本质主义问题并继续他们的工作，但那些问题成了公众理解心理学的障碍。当一个不了解情况的批评家声称心理学没有取得进步时，公众就会迷惑了。这类责难极少遇到挑战，这也反映了本书序言中所述的不幸事实：对于心理学领域所取得的科学成就及其意义，公众的了解是极度匮乏的。当我们仔细审视那些对心理学的批评，不难发现它们通常归结于一点：心理学至今没有为它提出的问题提供终极答案。对于这种指责，心理学毫不犹豫地低头认罪——像所有其他科学学科一样。 一些人可能会很不舒服地发现，包括心理学在内，没有任何科学可以对本质主义问题做出回答。霍尔顿和罗勒（Holton & Rolkr，1958）讨论过，当外行人被告知物理学不能够回答本质主义问题时所表现出的那种不安。他们谈论的是与放射性衰变有关的现象：发生衰变的放射性元素的原子数量与时间是呈指数函数关系的。可是，这种函数并不能解释为什么放射性衰变现象会发生。这个问题的解决将可能再次涉及另一个数学函数，但是它还是不能告诉外行人什么是真正的放射性衰变。霍尔顿和罗勒告诉我们：“我们必须平静地接受现代科学的局限性，它并没有声称可以发现'事物究竟是什么'”（pp.219-220）。科学作家罗伯特·怀特（Wright，1988）解释说： 同样，如果那些为人类本性问题寻求本质主义答案的人求诸于心理学，注定将会失望。心理学不是宗教，它是一个试图对所有行为做出科学解释的广阔领域。因此，心理学现在的解释是对行为的暂时性的理论建构，就目前来说，这些建构在解释行为方面优于其他解释。这些建构在将来注定会被更好的、更接近事实的理论概念体系所取代。 在评估一个心理学理论的可证伪性时，操作性定义的理念是一个非常有用的工具。概念有没有直接或间接地建立在可观测操作的基础上，是识别不可证伪的理论的重要线索。没有建立在可观测操作基础之上的概念，通常是为了拯救那些不能被数据印证的理论。所以，那些不严格的概念——理论学家不能为它们提供直接或间接的操作性联系——都应该引起怀疑。 与之相关的是科学家称之为“节省”的原则。节省原则是指，当两个理论有同样的解释能力的时候，较为简单的理论（涉及更少的概念和概念性关联）胜出。原因是，拥有较少概念性关联的理论在将来的检验中会更具可证伪性。 深刻理解操作主义的原则，也有助于我们识别不具备科学意义的问题。例如，在我的电脑文件夹里，有一篇来自于国际联合出版社的在线服务文章，标题为《动物会思考吗？ ".这篇文章讲述了动物行为方面最新的实验。文章中所引述的研究没有任何错误，但是，显然这个标题仅仅是一个玩笑。这个标题的问题在于没有科学意义，没有关于“思考”的操作性标准。许多报纸的标题中存在类似的问题，比如“计算机会思考吗？”没有操作性标准的话，这个问题也没有科学意义，但在鸡尾酒会上倒是可以大派用场。 实际上，观察人们在后面这个问题上的争论具有启发意义，因为这样的一个争论为我们亲眼见证先前讨论过的心理学中的“预设偏见”问题提供了机会。大部分人都有一个强烈的预设偏见，不希望计算机能够思考，为什么呢？出于各种原因，外行人认为“思考”这个概念与“人类”这个概念紧密联系，许多人在情感上不能接受非人类的物体（例如，计算机或看起来不像地球人类的外星生命形式）也能思考。 可是，尽管大部分人对会思考的计算机这一设想表示强烈反对，但他们并没有认真思考这一问题，也没有对“思考”做出一个更好的定义，使其能包含大部分人类（例如，婴儿）而排除所有计算机。有时，那些不熟悉人工智能进展的人提出的定义听起来颇为滑稽，因为他们总是选择了计算机能够做的事情作为标准。例如，许多人提出以“从经验中学习的能力”为标准，但一些计算机和人工智能已经达到这个标准了（Churchland，1995; Clark, 2001; Mc Corduck, 2004; Pfeifer & Scheier, 1999）。预设偏见的力量在这种情况下显而易见。会有人这样反应吗？“哦，我不知道。既然有些计算机符合我提出的'思考'的标准，那么我不得不说至少有些计算机是能够思考的吧！”通常大家是不会做出这样理智而诚实的反应的。更常见的情况是，人们开始寻找另一种标准，并期望计算机不能满足该标准。 通常人们的第二选择是“创造性”（“想出一些有用而且没有人想到过的东西”——我们先不管大部分人是否满足这个标准）。当被告知大多数专家都认可计算机能够达到这个标准时（Boden, 2003；Pfeifer & Scheier, 1999），人们仍然不愿承认机器思考的可能性。人们常常不会想到要做出一个操作性定义，转而提出计算机是不可能思考的，因为“是人类制造了计算机并且设计了程序；计算机只是执行程序而已”。 尽管这是反对机器思考的最古老的观点之一（Mc Corduck，2004, Robinson, 1992; Woo Uey, 2000），但它还是错误的。预设偏见让许多人认识不到，这些辩解与要讨论的问题毫不相干。几乎每一个人都认为思考是发生在自然世界中的一个过程。现在请注意，我们在讨论其他过程时，并不涉及“起源”的问题。考虑一下食物加热的过程。想想这个问题：“炉子能加热吗？”我们会说：“炉子不能加热，因为炉子是被人类制造出来的。因此只能说是'人'在加热。真正加热的不是炉子。”或者，什么是“举重”？起重机能“举重”吗？我们的答案是否仍旧为“起重机不会举重，因为起重机是由人造出来的。因此，我们只能说是'人'在举重。起重机真的不能举重”？当然不能这样说。一些事物的起源与它执行某一特定任务的能力是完全不相关的。思考的过程也是如此。一个事物能否思考并不依赖于这个事物的起源。 因为人们无法理性地接受机器有思考的能力，著名的计算机科学家艾伦·图灵（Alan Turing）设计出著名的“计算机能否思考”的实验。图灵设计的实验是一个操作性的实验，这对于我们的讨论是非常重要的。1950年，图灵在题为《计算机器与智能》的著名文章中写道，“我建议去考虑'机器可以思考吗？'这个问题。”他不想在鸡尾酒会那样的场合随意谈论这个问题，也不想如本质主义者那样无休止地讨论“思考”是什么意思，而是提出一个严格的操作性测试。他的想法是：如果计算机能够进行智能对话的话，那么就可以说它是有思考能力的。 图灵这一设想中的创造性在于，他提出一种方式将问题变得可操作化，同时又防止了“预设偏见”的干扰。图灵对于检验计算机是否可以进行智能对话的测试逻辑进行了严格的限定。这个测试并不是让测试者通过键盘和屏幕与计算机互动，然后由测试者判断计算机有没有进行智能对话。图灵没有采用这种设计，因为他很担心“预设偏见”的干扰。图灵确信，一旦一个人坐到计算机、键盘和屏幕（显然是一些机器）之前，无论这个机器做什么，这个人都会否认它有思考能力。因此，图灵提出应控制与思考能力无关的外在因素。其著名的实验设计是让测试者通过两个键盘对话（一个和计算机相连，另一个和人相连，并且都在视线之外），然后再判断哪个是人，哪个是机器。如果被试不能以大于随机水平的正确率猜出哪一个是人，那么我们就有理由推断计算机具有同人一样的对话能力，而对话能力正是“思考”的操作性定义。 图灵的主要思路“与交响乐团选拔乐师的试听面试的思路是一样的，试听面试时，在评委与面试乐师之间放置一个不透明的屏幕，前者要隔着这个屏幕来判断乐师演奏的好坏。很显然，评委关心的是音乐能力，而且也仅仅是音乐能力而已。性别、头发的长度、皮肤颜色和体重等都是完全不相关的……图灵认为，人们对智力的判断可能同样受到对方是否拥有柔软的皮肤、温热的血液、面部特征、手和眼睛等那些明显不是智力本质的因素的影响”。（Dennett, 1998, p.5）图灵的测试启发我们，如果我们想要理性地讨论心理学概念，那么操作性定义是必需的；我们要以一种有条理的方式进行判断，而不仅仅根据我们自己对某个问题的偏见。 观察人们讨论人工智能问题时所展示的思维方式，就能发现科学和非科学思维方式之间的区别。科学的方式是先发展一种合理的操作性定义，然后看我们可以从中得出哪些关于思考、计算机和人类的结论。与之不同的是，预设偏见主导了大多数人的思维。他们已经得出了某一结论，并且对于计算机和人类的表现之间已被发现的差异并不感兴趣。相反，一旦形成定势，他们就会绞尽脑汁地去找出各种理由，来巩固自己的这些想法，避免发生改变。于是，我们看到，正是预设偏见和非操作性的本质主义态度，让人们认定他们“就是知道”思维到底是怎么回事。这种态度使大多数人的直觉心理理论无法证伪，因此完全无用。也正是这种态度说明了为什么我们需要科学心理学！ 操作性定义是利用可测量、可观察的操作来表述的概念定义。我们确信某个理论具有可证伪性的主要途径之一，就是确定理论中的关键概念具备可用可重复性很强的行为观察来表述的操作性定义。操作性定义是让科学知识变得公开可检验的主要机制。这样的定义被置于公共领域，使其所界定的理论性概念能够接受所有人的检验，而不是像“直觉的”、非经验性的定义那样，只属于特定个体，检验它的机会并不向所有人开放。 由于心理学使用一些来源于日常生活的词语，如智力和焦虑，许多人对于这些术语的含义有着预设的想法，因此往往意识不到对这些术语进行操作性定义的必要性。心理学和所有其他科学门类一样，也需要对其术语进行操作性定义。可是，人们常常要求心理学家回答本质主义的问题（有关概念的纯粹深层本质的问题），而其他科学家就不必回答这类问题。没有科学能够回答这样的终极问题。心理学和其他科学门类一样，正在试图不断地完善其操作性定义，使理论概念能够更加准确地反映真实世界的原貌。