Why Stuttering Experts Don't Agree
- Cognitive Biases
- Well-Structured and Loosely Structured Fields
- Observable and Unobservable Phenomena
- Scientific Revolutions
This article is best read after reading the Three Eras of Stuttering Treatment.
A 2005 study of a stuttering modification therapy (Van Riper therapy) program found that immediately post-treatment subjects' speech improved on average 10%. Six months later this small gain had all but disappeared. Several measures of anxiety found 10-15% psychological improvements, after six months. The researchers concluded that the program "…appears to be ineffective in producing durable improvements in stuttering behaviors." [1]
Three speech-language pathologists wrote letters to the journal objecting to aspects of the study. Bruce Ryan wrote:
As to the goal of treatment, I have to respectfully disagree with the authors' statement that "…reduced frequency of stuttering was not an overt goal of the [therapy program]"…Fluency, or at least very good management of stuttering, was one of Van Riper's main goals to be attained in- and out-of-clinic. This is what he advocated when I was with him as a student in 1956-1957 and it is what he had written in his books. Fluency is at least one goal of the SSMP as I read the manual, or why the use of prolongations, cancellations, and so on to modify the speech? From [the book The Treatment of Stuttering, by Charles] Van Riper, "We tell [the person who stutters] that some stutterers achieve complete fluency, at least equal to that characteristic of most normal speakers." [2]
Peter Reitzes and Gregory Snyder wrote:
we argue that [the efficacy study] may be inappropriate relative to the stated therapeutic goals…Specifically, these goals include approaching stuttering in nonavoidant ways and using stuttering modification strategies to move forward through moments of stuttering. As a result, a valid and important therapeutic objective for some clients who participate in…stuttering modification approaches may include increasing overt stuttering frequency as a means of desensitization to the fear of stuttering and reducing covert stuttering behaviors. As [Walter] Manning has written, "Increased stuttering usually occurs with decreased avoidance. So, under certain circumstances, one sign of [therapeutic] progress could very well be an increase in the frequency of stuttering." Consequently, the use of stuttering frequency as an indicator of treatment efficacy may be an inappropriate and insensitive measurement relative to the treatment efficacy of…stuttering modification based therapies. [3]
Dividing stuttering treatments into three eras is an exercise in hindsight bias. To make neat categories of treatments I'm emphasizing some facts and ignoring others. The goal of Travis, Bryngelson, Johnson, and Van Riper was fluent speech. They tried to develop a neurological treatment based on "confused or ambiguous lateral dominance," [4] e.g. forcing left-handed stutterers to use their right hands. This didn't work. They tried to train stutterers to talk fluently, using cancellations, pull-outs, etc. This was also unsuccessful. They tried to reduce stutterers' speech-related fears and anxieties, and were successful. Today, practitioners of stuttering modification therapy, such as Reitzes and Snyder, argue that fluent speech is not the goal of their therapy, but rather their goal is to reduce speech-related fears and anxieties and improve communication.
Cognitive Biases
Cognitive biases are mental shortcuts (or heuristics) we use to make decisions or to persuade other people to agree with our point of view, when we lack complete information or when we have too much information and lack the time to examine all of it. More than one hundred cognitive biases have been identified. [5]
Cognitive biases can help us reach satisfactory conclusions, e.g., buying the same make of car that has satisfied you in the past. But cognitive biases can also lead to wrong conclusions and bad decisions. For example, my first stuttering treatment was a prolonged speech (fluency shaping) therapy program in 1981, when I was nineteen years old. I learned to speak slowly but fluently in the speech clinic. I was unable to use this fluent speech in conversations outside the speech clinic (the therapy failed at the transfer stage). In 1984 I did a "refresher" of the same program and this time was able to speak fluently outside of the speech clinic—for three days. Then my stuttering returned.
In 1990 I tried speech therapy again, with a recent Ph.D. speech-language pathologist who'd written her dissertation on stuttering. Twice a week she spent an hour telling me that adult stutterers with my severity can never talk fluently, and that I must accept that I would be a stutterer for the rest of my life. I objected that this wasn't true, that I'd learned to talk fluently at another speech clinic. She kept telling me to change my goals to accommodate stuttering. I quit seeing her after six weeks.
Confirmation bias is the interpretation of information to support one's preconceived ideas. From her point of view, I'd done two prolonged speech stuttering therapy programs and I still stuttered, proving her belief that prolonged speech stuttering therapy doesn't work. From her point of view, I'd been taught some tricks that produced fluency only in the speech clinic, or temporarily outside the speech clinic. To her, I was proof that the core behaviors of stuttering are immutable. To her, my only hope was to change my attitude and my goals and learn to communicate despite stuttering. To her, the Iowa therapies were truth and prolonged speech was smoke and mirrors.
From my point of view, I was certain that I could talk fluently. I'd done it twice, and I knew I could do it again, if I could find a speech-language pathologist to teach me. I didn't want to accept stuttering. I'd didn't want to b-b-b-bounce through v-v-v-voluntary st-st-st-stuttering. I didn't want to watch video tapes of my stuttering, or listen to support group debates about whether we were "stutterers" or "persons who stutter" or "persons who sometimes stutter." [6] I believed that prolonged speech therapy was the modern method and stuttering modification therapy was out of date and ineffective.
Another type of cognitive bias is ingroup bias or the bandwagon effect. This cognitive bias occurs when an individual makes a decision (e.g., recommending a stuttering treatment) based on what other people do, especially what members of one's group do.
Experts studying misguided [medical] care have recently concluded that the majority of errors are due to flaws in physician thinking, not technical mistakes. In one study of misdiagnoses that caused serious harm to patients, some 80 percent could be accounted for by a cascade of cognitive errors…Another study of one hundred incorrect diagnoses found that inadequate medical knowledge was the reason for error in only four instances. The doctors didn't stumble because of their ignorance of clinical facts; rather, they missed diagnoses because they fell into cognitive traps. [7]
Cognitive biases are why intelligent, educated, respected experts can't agree. Experts make mistakes, such as recommending ineffective treatments, not because not enough studies have been done or because not enough books have been written. Experts fail because cognitive biases skew their thinking.
Well-Structured and Loosely Structured Fields
Well-structured fields have observable phenomena, enabling anyone to judge whether a hypothesis is correct. [8] Examples include math, physics, and computer science. Well-structured fields have clearly defined terms and rules that everyone agrees on (e.g., the laws of physics). Persons in these fields easily communicate with each other, and with persons in other well-structured fields (e.g., physicists worked with astronomers to test Einstein's theories). In well-structured fields young people quickly learn the rules and are rewarded for new ideas. [9]
Loosely structured fields lack observable phenomena. Examples include religion and psychology. Loosely structured fields have mythology, traditions, and rituals ("this is the way we've always done it") instead of laws. Authoritative experts interpret observed phenomena in terms of myths. E.g., some ancient Chinese believed that eclipses were the celestial dragon eating the Sun or the Moon, and people had to bang drums and pots to scare the dragon away. [10]
In loosely structured fields, each "school of thought" has its own definitions of terms, and each school's data isn't recognizable or quantifiable by other schools. Each "school of thought" has its own values and goals. E.g., in the exchange of letters between Ryan, Reitzes, and Snyder you saw that stuttering modification therapy has been practiced for more than fifty years, yet speech-language pathologists can't agree what the goal of the therapy is: fluent speech; or non-avoidance of stuttering, including increased or voluntary stuttering as "strategies to move forward through moments of stuttering"?
Evidence-based practice (EBP) integrates scientific research, clinical decision-making, and patient outcomes. [11] Clinicians are expected to provide treatments that have been proven effective via analysis of observable data. But EBP has been slow to catch on in stuttering therapy:
the evidence base in fluency disorders is difficult and controversial [with] a paucity of well-controlled assessment and treatment studies. Contributing to this problem is lack of agreement on outcome measures, varied definitions of success and the heterogeneity of our clients. [12]
Members of each "school of thought" don't read papers written by members of other "schools of thought." It's not simply ignoring differing points of view: cognitive biases make members of different "schools of thought" unable to understand each other.
Young people in a loosely structured field are expected to apprentice to an older mentor. Young people are rewarded for parroting old ideas. New ideas are encouraged only if they are repackaged old ideas. New ideas outside of the "school of thought" are ignored.
Observable and Unobservable Phenomena
The field of stuttering is a mixture of observable and unobservable phenomena. Some areas are well-structured and other areas are loosely structured.
The Iowa era was loosely structured. At the University of Iowa in the late 1920s Bryngelson and others tried to study the neurology of stuttering, but the methods available at the time (e.g., switching left-handed people to use their right hands) were too crude to make adequate observations. [13] Johnson observed young children soon after the onset of stuttering but then ignored his observations when the data conflicted with his diagnosogenic theory [14] (which wasn't a theory at all, but rather was a myth). Van Riper's stuttering modification therapy aimed to reduce adult stutterers' unobservable speech-related fears and anxieties, and to modify stuttering in ways that were difficult to measure, e.g., reducing the severity of disfluencies while using voluntary stuttering to increase the number of disfluencies.
The prolonged speech stuttering therapy era focused on observable phenomena: speech disfluencies. However, different researchers measure disfluencies in different ways, e.g., disfluencies per word vs. disfluencies per syllable. Some researchers use statistics to analyze their data, when others don't [15] (statistics make unobservable effects observable, but only to people who understand statistics).
Iowa era practitioners described treatment efficacy as if there were sub-types of stutterers, [16] e.g., Van Riper's "four tracks" of stutterers, and studies were reported as percentage of subjects helped (e.g., a treatment was effective for half the stutterers [17]). In contrast, prolonged speech practitioners see stuttering as one disorder, so studies provide results as averages of all subjects in study. Neither method is necessarily better than the other, but it makes comparisons between stuttering modification and prolonged speech programs difficult. For example, the National Stuttering Association position paper about anti-stuttering devices states:
there are no published, independent studies that show what percentage of the population of people who stutter are likely to benefit from the SpeechEasy. [18]
The statement is true, but several studies provide results averaged across the subjects. Each school of thought's data isn't recognizable or quantifiable by others.
In the neurological era brain imaging enables us to observe previously unobservable phenomena, such as auditory processing activity. But brain scans are far from ideal. Most researchers don't have access to a brain scanner, and even if they did they wouldn't know how to interpret the pictures.
It's an alluring prospect, but the approach is still viewed with suspicion by mainstream psychiatrists…journals churn out hundreds of brain imaging articles each month [but] we haven't quite figured out what these pictures mean. Are we really seeing the mind in action, or are we allowing ourselves to be seduced by images that may actually tell us very little?…In recent years, functional neuroimaging research has yielded a wealth of intriguing fodder for journalists but few scientific breakthroughs…
Yale researchers gave participants various nonsensical explanations of human behavior. Half the time, the researchers added the phrase "Brain scans indicate" before the explanation, and then inserted the spurious finding. When the brain-speak was added, participants judged the explanations more satisfying. [19]
Scientific Revolutions
Thomas Kuhn, in his 1962 book The Structure of Scientific Revolutions, coined the term "paradigm shift" to mean a change in basic assumptions in a field of knowledge. The change from the Iowa era to the prolonged speech era was a paradigm shift: the assumption that stutterers can never learn to talk fluently changed. Another paradigm shift occurred in the 1990s, changing the assumption that the neurological causes of stuttering can't be changed.
But the history of stuttering treatments raises another question: perhaps paradigm shifts are seen only with hindsight bias? E.g., the Iowa researchers tried but failed to develop neurological treatments and to train stutterers to talk fluently, succeeded in treating stutterers' speech-related fears and anxieties, and contemporary speech-language pathologists use hindsight bias to create a paradigm that we call the "Iowa therapies" that the Iowa researchers never believed.
Kuhn also coined the terms normal science and revolutionary science. Normal science agrees on a paradigm or set of basic assumptions. Kuhn wrote that normal scientific work is akin to "puzzle-solving" in the sense that puzzles always have an answer. Scientists work on problems expecting to find single, clear answers.
Kuhn noted that every scientific field has anomalies that are difficult to explain within accepted paradigms. Some scientists are bold while others are conservative. The bold scientists propose a new paradigm, in a process Kuhn called revolutionary science, and the conservative scientists oppose the new paradigm.
Kuhn also noted that different scientific paradigms aren't comparable, or are incommensurable. Assumptions are different, terminology has different definitions, and what questions are valid are different. Scientists can't rationally compare one paradigm to another, and choose the one that best explains the facts. The tools and evidence used to support each paradigm are developed within the paradigm, so supporters of a paradigm believe they have proof that their paradigm is correct. Kuhn wrote:
Though each may hope to convert the other to his way of seeing science and its problems, neither may hope to prove his case. The competition between paradigms is not the sort of battle that can be resolved by proof. [20]
Kuhn's ideas are controversial, especially the idea that scientific methods can't help intelligent, educated, respected experts to agree as to what ideas are right and what ideas are wrong.
Like all human activities, fields of science are rooted in their social and cultural contexts. If you can sort the apples from the oranges of stuttering treatment you'll be better able to make informed choices as to the best treatment for yourself or your child.
References
[1] Blomgren, M., Roy, N., Callister, T., & Merrill, R. (2005). Intensive Stuttering Modification Therapy: A Multidimensional Assessment of Treatment Outcomes. Journal of Speech and Hearing Research, 48, 509-523.
[2] Ryan, Bruce. (2006). Response to Blomgren, Roy, Callister, and Merrill (2005). Journal of Speech, Language, and Hearing Research, 49, 1412–1414.
[3] Reitzes, P., & Snyder, G. (2006). Response to "Intensive Stuttering Modification Therapy: A Multidimensional Assessment of Treatment Outcomes," by Blomgren, Roy, Callister, and Merrill (2005). Journal of Speech, Language, and Hearing Research, 49, 1420–1422.
[4] Bloodstein, O., & Bernstein Ratner, N. (2007). A Handbook on Stuttering, Sixth Edition. Clifton Park, NY: Thomson. ISBN 978-1-4180-4203-5, pages 116–120.
[5] Wikipedia, "List of cognitive biases."
[6] Ham, Richard. (1999). Clinical Management of Stuttering in Older Children and Adults. ISBN 978-0834211179.
[7] Groopman, Jerome. (2007). How Doctors Think, page 24.
[8] Csikszentmihaly, Mihaly. (1996). Creativity: Flow and the Psychology of Discovery and Invention (New York: HarperCollins).
[9] The five youngest Nobel Laureates were all in Physics, with ages ranging from 25 to 31. In contrast, the youngest Nobel Laureate in Economics was 51. No speech-language pathologist has won the Nobel Prize in Physiology or Medicine. http://nobelprize.org/contact/faq/index.html#3b
[10] Solar Eclipses in History and Mythology, http://www.bibalex.org/eclipse2006/HistoricalObservationsofSolarEclipses.htm, accessed 2008 May 17.
[11] Banotai, Alyssa. (2005, Nov 14). Emphasis on Evidence. ADVANCE for Speech-Language Pathologists & Audiologists, 15, 7.
[12] Kully, D., & M. Langevin. (2005, Oct 18). Evidence-Based Practice in Fluency Disorders. ASHA Leader, 10, 10-23.
[13] Bloodstein, O., & Bernstein Ratner, N. (2007). A Handbook on Stuttering, Sixth Edition. Clifton Park, NY: Thomson. ISBN 978-1-4180-4203-5, page 116.
[14] Bloodstein, O., & Bernstein Ratner, N. (2007). A Handbook on Stuttering, Sixth Edition. Clifton Park, NY: Thomson. ISBN 978-1-4180-4203-5, page 308.
[15] Ingham, R., Moglia, R., Frank, P., Costello-Ingham, J., & Cordes, A. (1997). Experimental Investigation of the Effects of Frequency-Altered Auditory Feedback on the Speech of Adults Who Stutter. Journal of Speech, Language, and Hearing Research, 40, 361-372. Kalinowski, J., Rastatter, M., & Stuart, A. (1998). Altered Auditory Feedback Research Conditions and Situations of Everyday Life: Comments on Ingham, Moglia, Frank, Costello Ingham, and Cordes (1997). Journal of Speech, Language, and Hearing Research, 41, 511-513. Ingham, R., Moglia, R., Frank, P., Costello-Ingham, J., & Cordes, A. (1998). The Effects of Frequency-Altered Feedback on Stuttering: Reply to Kalinowski, Rastatter, and Stuart (1998). Journal of Speech, Language, and Hearing Research, 41, 513-515.
[16] Guitar, Barry. (2005). Stuttering: An Integrated Approach to Its Nature and Treatment (3rd Edition) ISBN 978-0-7817-3920-7, page 19.
[17] Bloodstein, O., & Bernstein Ratner, N. (2007). A Handbook on Stuttering, Sixth Edition. Clifton Park, NY: Thomson. ISBN 978-1-4180-4203-5, page 352.
[18] Yaruss, S. Gabel, R. "The National Stuttering Association’s Position on the Speech Easy and other assistive devices," http://www.nsastutter.org/material/index.php?matid=328, dated February 16, 2005.
[19] Carlet, D. (2008). Mind Readers. Wired, 16, 122.
[20] Kuhn, Thomas. The Structure of Scientific Revolutions, page 148.
