“TESSITURA”1 IS A term for a singer’s most comfortable vocal range—the range in which she will make her most beautiful music. A mezzo-soprano can hit a high C, but she has to strain to stay there, and you can hear it. Tessitura is the place where you find your true voice, where you can have a long and healthy career. The quest for my tessitura in psychology was a search in three dimensions: style, tempo, and content.
“EVERY TIME I construct a sentence,” said Dick Solomon to Steve and me, “I first talk to the person perched on my right shoulder.” The occasion was dinner at my favorite restaurant, Le Bec Fin. Georges Perrier had created this temple of classic French cuisine in the late 1960s, and I became one of his poorest but steadiest customers. I became comfortable there. I saved up, and whenever there was anything to celebrate, I visited. On this evening the dish was the lightest quenelle de brochet imaginable in a coral-hued lobster sauce, and the wine was a pale yellow Puligny Montrachet. Dick, Steve, and I were celebrating the acceptance of a paper by a good journal.
“On my right shoulder sits Harold Schlosberg, my mentor at Brown,” continued Dick. “I read the sentence to Harold very slowly, and if he nods his head, I turn to my left shoulder. On my left shoulder sits my other mentor, Walter S. Hunter. I now read him the sentence very slowly, and if he also nods, the sentence goes in. If not, I start again.”
Dick transported his graduate mentors at Brown in his imagination when he went off to teach at Harvard, and he then brought them to Penn. I didn’t have to imagine my mentors, because they were here with me in Philadelphia. Two of them, Dick and Frank Irwin, would serve my journal style well by perching on my shoulders since each edited a major journal. I didn’t need to imagine their comments, since I routinely sent each of them an early draft of any paper I was writing.
Dick and Frank were very careful in print, and so they were the ideal editors of the Psychological Review and the Journal of Experimental Psychology, respectively. In their writing they crafted long sentences, and any one lifted out of context was self-contained and likely true.
Here’s a typical one from Dick:
Because we shall show that discriminative instrumental acts can be established without those acts having been reinforced differentially (because they could not occur under curare), then a revision of theories emphasizing the importance of peripheral, skeletal responding and proprioceptive cues seems definitely to be required.2
And a typical one from Frank:
The present experiment was designed to determine whether, as predicted on these grounds, the number of cards required for decisions in the expanded judgment situation would increase with the value of money prizes for correct decisions and would decrease with increased money cost per card.3
How many times did you have to read each sentence to understand it? Twice for me, and I know these topics well. Their style was “reviewer friendly,” and indeed Dick and Frank rarely encountered a journal rejection, but their style was not “reader friendly.”
I tried not to write like they did.
Most readers are busy. I sure am. Many readers just scan. I know I do. Help the scanners by using short sentences and plain words. The big word and the long sentence must increase accuracy a lot to make up for impeded reading. And don’t overwrite. Omit extra words and information that the reader already knows. Overwriting slows the reader way down and does not increase accuracy much at all.
In style, even if it defied the American Psychological Association’s best-selling publication manual,4 I had found my tessitura.
MY TEMPO WAS wrong. My father, Adrian, was fast, and Irene said that he could do in one hour what other lawyers took eight to do. But Irene also said that he then could use the extra seven hours to check and perfect. I had Adrian’s speed, but I just raced on to the next task rather than checking and perfecting. My speeding was not limited to work, and I suspect that my poor tempo in part led to others’ finding me brusque and inferring that I was self-absorbed.
Are there any slow psychologists? There was at least one, William K. Estes, the premier mathematical learning theorist. Conversing with Bill was agony.
“Bill,” I asked at a party where he was the guest of honor, “what do you think the evolutionary function of sleeping is?” He stared at me unblinking. I waited. I waited. It was so long I started to count to myself. At forty-five, he blinked and his lips finally moved.
“What, Marty, do you think is the evolutionary function of being awake?”
Could I learn to be slow, not as slow as Bill Estes, but slow enough not to offend?
THE DEPARTMENT WAS hiring someone who worked in vision, the most scientifically sophisticated branch of psychology. Ed Pugh, the candidate, gave his job talk on how many photons of light it took to fire a single retinal cell. Between one and seven seemed to be the answer. This was all way over my head, but at the reception afterward, I managed to ask Ed a question that was not completely idiotic. Ed waited a full ten seconds and then gave me a full answer delivered at the slow pace I needed to understand him. He also made me feel that I’d asked him a first-rate question that took him to the cutting edge of visual science.
“You’re really slow, Ed,” I commented and immediately explained that I meant it as a compliment. “Have you always been so slow?”
“I was an all-state quarterback in my Louisiana high school. There are no slow quarterbacks,” Ed replied. “I had to learn slow. After college I tried to become a Jesuit. Jesuit students have two mentors. One grades you, and the other is your ‘socius.’ Day after day my socius would write just one sentence and then instruct me to go sit under the tree in the cloister for the rest of the afternoon and think about that sentence.”
“Can you teach me to be slow, Ed?” I asked. After we hired Ed, he took me under his wing and taught me how to be slow. For a year we met once a week in my office to read Kierkegaard’s Fear and Trembling. We did only one page each week. I became slower.
Maybe too slow.
Alan Kors was a new assistant professor of history, fresh from his Harvard doctorate on Baron d’Holbach and the French Encyclopedists. Alan and I had been good friends for a decade. He was a classmate at Princeton, editor of our literary magazine, and the most entertaining of the undergraduate literati at Wilson Lodge. That he was an incomparable storyteller and improvised endless rhyming doggerel on his guitar at parties did not mask that he was also a serious poet. He was an outspoken atheist and, as best man at my wedding with Kerry, stood shoulder to shoulder with me against my mother-in-law-to-be, who cursed and wailed when I refused to allow the G word at the ceremony. It was a tremendous personal boon to me that he joined Penn’s history department.
“Let’s teach together, Marty,” suggested Alan. We shared many mutual interests: brainwashing, religious conversion, poetry, and the history of science. I agreed instantly. Alan was also as mesmerizing a teacher as he was a conversationalist, the only Penn professor I knew of to ever get perfect teaching ratings in a huge lecture course—a 5.0 from every single student.
I tried out my new slow, noninterrupting style with Alan. I sat in the back row of the class we were to teach together. Alan began his lecture. About one hundred undergraduates were present, and the topic was science during the Enlightenment. At several points, he brought up issues that I knew a bit about, and I was tempted to interrupt. But, no, I waited for Alan to introduce me. The lecture went on … and on. It was fascinating and entertaining, pure Alan Kors. But he did not introduce me. After ninety minutes, the lecture was over, and I had been silent. Alan and I walked back to the student center for coffee.
“What happened?” I asked, miffed. “You were supposed to introduce me.”
“What happened?” Alan replied, also miffed. “You were supposed to interrupt—like you usually do.”
My tempo was far from perfect, but it was slowing down.
CONTENT WAS THE final element of my tessitura. Behaviorism at its heart was trying to banish two things from science: mental life, of course, and the future—teleology, forward-looking explanations of action. These issues played out in my own department, often in tense conversations. One occurred in 1972 in the “chance encounter” room, the windowless, grotty basement of one of the three buildings among which the bursting-at-the-seams psychology department was scattered miserably. In the hope that some psychologists would bump into each other occasionally, a handful of us made a practice of lounging there, drinking instant coffee. Dave Williams popped in often, and we continued our endless argument over whether people and animals have a rich mental life, with Dave saying they don’t and me saying they do, but I could sense that Dave was giving in and abandoning the barren world of Skinnerian psychology. He was on a new but flailing search to be someplace lusher and greener.
Even though smoking was forbidden, Frank Irwin spent a lot of time in the chance encounter room. This was his last year of teaching, and he remained the most upright pillar in the most rigid of psychology departments. All the hours he spent there were emblematic of what an upright departmental citizen he was.
I told Frank that Jim Johnston and I were slaving away on a comprehensive cognitive—“Irwinian”—theory of avoidance learning. Our theory was a stark contrast—some might say a “slap in the face”—to the pursuit to which Neal Miller and Dick Solomon had both devoted their past two decades. If correct, our theory would not make me any more popular with my fellow learning theorists, who already saw me as suspect because of the cognitive slant of learned helplessness.
Learning how to avoid something unpleasant had become the testing ground of the entire behaviorist program. Neal and Dick wanted to show that avoidance could be completely explained without any reference to either the future or cognition. The future cannot act on the present, they reasoned, so any teleological explanation has causality working backward in time, which is a philosophical error of the first water.
In avoidance learning, animals were first given escape training in which a tone preceded a shock by five seconds. The animals could escape the shock by jumping over a barrier to the safe side. Jumping over the barrier also turned off the signaling tone. Animals quickly learned to jump as soon as the tone went on, but before the shock. This was called avoidance learning because the animals avoided more shock by jumping as soon as they heard the tone—which both prevented the shock and terminated the tone. Avoidance, to the uninitiated, looked like a prime case of a future-directed cognition, the lay explanation being If I jump, I will avoid future shock. So it became a crucial task of the behaviorists to show that they could completely explain the jumping without reference to the future. This was quite a trick, and an explanation of avoidance referencing only the present and the past was considered a tour de force for learning theory.5
Here’s how the explanation—two-process theory—worked. It denied that the animals expected anything at all, claiming that a nonoccurring event—the avoided shock—could have no role in conditioning. Instead, a present event, the tone coming on, became fear-evoking by the usual process of Pavlovian conditioning, having previously been paired with shock. So jumping, by getting the fearful tone to stop, was actually instrumentally reinforced and not at all motivated by avoidance of a shock that never came on.
By contrast, Jim and I claimed that the animals learned first that the tone predicted shock and then that jumping during the tone, but before the shock went on, was followed by no shock. Thus they learned the expectation If I jump during the tone, I will get no shock.6 That expectation—an if-then conditional about a future outcome that does not happen—explained avoidance.
This was a rare instance of a head-on collision between two theories where a crucial prediction could settle the argument, in this case, about what happens during extinction. Once the animals became steady jumpers, they never got shocked again, and so the tone was no longer paired with shock. This “Pavlovian extinction” should have caused the tone to lose its fearful properties. If the animals were jumping merely to turn off the fearful tone, as two-process theory contended, they should stop after enough trials to extinguish fear of the tone. If, on the other hand, they were jumping to prevent the anticipated shock, as Jim and I contended, the jumping should continue, as every jump would bear out their if-then expectation that jumping would prevent shock—extinction breaks the association of tone and shock, but it does not disconfirm the conditional If I jump, I will not be shocked. In fact, every “extinction” trial continued to confirm the conditional expectation. So if this expectation underpinned the jumping, jumping would persist.
Which is exactly what happened. Hundreds of trials later, the animals were still jumping. They also looked nonchalant, not like animals escaping fear.7
Jim was obsessive, and it took us more than a year of arguing, sentence by sentence, with each other to work out every detail. We went through all the evidence exhaustively; it was entirely consistent with a cognitive theory of avoidance and inconsistent at the crucial points with the dominant two-process theory. Once we were sure, I sent Frank Irwin a draft to follow up our chance encounter room chats and presented a colloquium to the department. Frank had had a very long dry spell, and this was the first and only major use ever made of his theory. He sent me a handwritten letter saying how happy he was to have lived to see his way of looking at learning come to fruition. I was pleased this time to receive such a fulsome thank-you, rather than his usual lake of red ink.
This paper was a major “teaching moment” for me. I was sure we’d nailed it and that two-factor theorists would acknowledge that we were right.
But I was under a misapprehension about confrontation in science. I thought that science should proceed by looking for crucial differences between theories and then testing the opposing predictions. This is what Carl Sagan called “good” science: the moon either is or is not covered with a thick layer of dust, and when humans landed on the moon, the dust theory of the origin of the solar system was proved wrong and abandoned. By 1973, I had been in the ring for three heavyweight boxing matches with the dominant stimulus-response-reinforcement theorists.
The first was over learned helplessness. In response to our contention that learned helplessness entailed the cognition Nothing I do matters, learning theorists had posited that a few random shock-offsets had just superstitiously reinforced a motor response of standing still. Steve Maier killed that one by explicitly reinforcing standing still in a group of dogs and showing that they were not helpless. I suppose I expected the Skinnerians to stand up and concede that they were wrong and we were right, but they did not. Rather, learning theorists made almost no response and thereafter just went silent about learned helplessness. They simply excluded it from the learning theory corpus thereafter.
Then there was John Garcia and sauce béarnaise. I suppose I expected the Pavlovians, who held the premise that any conditional stimulus paired with any unconditional stimulus would result in conditioning, to hang it up and admit they were wrong. But again they were almost silent—uncomfortably so about Garcia, although they grudgingly elected him to the National Academy of Science.
Finally, there was the cognitive theory of avoidance. I was surprised by the reaction from our learning theory colleagues—there was none. Dick Solomon, perfect gentleman that he was, never said a word to me about this paper. Neal Miller, Bob Rescorla, and the rest of the learning theory gang stayed silent and stayed silent and stayed silent.
Instead of an abject surrender, there was no surrender at all, just a gradual drift. I should have known. Behaviorist learning theory eventually relinquished its ambition to explain “all behavior.”8 It abandoned wholesale anti-mentalism and subtly allowed in “expectations” to explain conditioned behavior. The picture of conditioning itself underwent gradual revision, resulting in the view that “the organism is better seen as an information seeker using logical and perceptual relations among stimuli, along with its own preconceptions, to form a sophisticated representation of the world.” Thus the great program of learning theory, the attempt to explain behavior without positing the “forbidden” mental representation of possible futures, slowly folded. Even more importantly, the possibility of a science that takes seriously thinking about the future opened.
But why were thoughts about the future “forbidden” in the first place? A huge conceptual error animated behaviorism, in which something genuinely suspect—causation backward in time, the future determining the present—was conflated with something not at all mysterious, namely, guidance by here-and-now cognitions that bear information about possible futures. There is nothing mysterious, or causally backward, about present action being guided by present thoughts of what will happen in the future.
I also suppose it should have come as no surprise that my Skinnerian and Pavlovian colleagues were not grateful to me for my part in showing them the error of their ways. Rather, to be candid, I left a bad taste in their mouths.
And my takeaway? I too had a bad taste in my mouth, but I’d come to know my tessitura, my thinking’s comfort zone: I looked for the most basic premise of the prevailing theory and asked what would follow if it were entirely wrong? I was learning not to expect popularity, but in my heart I knew that truth would out.
“WHAT IS A ‘trade’ book?” I asked. I’d never heard of such a beast.
“It is smack in between a scholarly monograph, like you usually write, and a mass market book that you find in airports,” answered Buck (Hayward) Rogers. Buck, representing W. H. Freeman publishers, took me to dinner at Le Bec Fin. “You’ve been writing about helplessness. This is not a dry topic only for professionals. It’s of great interest to the educated public. Helplessness, you yourself have written, bears on depression, on death, and on raising kids. There has been only one such a ‘trade’ book in psychology, the Social Animal. Yours would be the second.”
I took the offer seriously. Such a project would require me to organize my thoughts on external validity—whether helplessness really matters—as well as give me an opportunity to justify and explain experimental psychology to the public. It seemed like the first step on a path to the unknown. Jim Johnston, with his usual skepticism, asked me if the world needed a whole book about helplessness.
BY 1974 I had found my tessitura.
In style, I was comfortable writing plainly: few big words, short sentences, brief paragraphs, and straightforward arguments.
In tempo, I was becoming slower. I was more deliberate in speech as well as thought. My pace had improved. Though imperfect it was closer to right than it had ever been.
Content is what really matters. Physics and chemistry had engineering before they became “basic” sciences. Physics predicted eclipses and tides and the trajectories of cannonballs. Alchemy invented medicines and paints, even if it could not make gold out of lead. Basic physicists and chemists therefore knew what merited further drilling down. Psychology, in contrast, had never had engineering—it had no discoveries and inventions that worked in the real world—so what was “basic” to psychology was arbitrary. This was what bothered me so about learning theory, because it was far from evident that Pavlovian conditioning and instrumental training were basic to anything.
My tessitura followed. I would choose psychological topics anchored in the real world, document them at the real-world level, and then drill down to find their elements.
The world did indeed need a book about helplessness.