Hank Pennypacker had an inestimable influence on the development of the science of behavior; the establishment and promotion of behavior analysis as a cohesive field and profession; the foundations, health, and vigor of the behavior analysis program at the University of Florida; and the achievements of the many students and colleagues whose lives and careers he inspired, including my own. I had the great good fortune to be part of Hank's research and development team while a graduate student during the early to mid-1980's, as the analyses and syntheses that would give rise to MammaCare were unfolding. This period arguably merits a closer look in our honoring of Hank, for it represents a genuine microcosm of his legacy as a scholar, a teacher, and as a person.

Hank often emphasized the power of one's initial experimental question in shaping a program of study. For this particular program, the origin story lay in the observation that a majority of breast tumors were being found first by women themselves rather than through screening technologies such as mammograms or ultrasound. Sadly, by the time the tumors were identified, they were often large and, if cancerous, advanced. Hank's question thus emerged: If fingers could become proficient with the complex discriminations required to read Braille, shouldn't it be possible to help women learn to detect small masses, at a point when treatment was much more effective? For the next 50+ years, Hank accepted the challenge this question entailed. To begin to address it, he turned first to his roots in classic experimental psychology. (Hank's PhD work involved highly controlled laboratory investigations of classical conditioning in monkeys, and he was known to describe himself as a “converted Hullian.”). Here, a series of careful psychophysical studies detailed that the tiniest of ball bearings embedded between silicon pancakes could be reliably reported when fingers were placed directly over the target (e.g., Adams et al., 1976), clearly demonstrating the proof of concept.

Of course, ball bearings are not breast tumors, and the need for a more realistic simulation of breast tissues resulted in an ongoing collaboration with materials scientists to develop a medium appropriate for discrimination testing and training. Hank frequently argued for the importance of working together with other disciplines when attacking big problems, and here was an initial case in point for the breast-examination project. With his legendary insistence on absolute units, critical dimensions of real breast tumors and normal nodularity were measured and approximated in silicone in an iterative process that resulted in increasingly sophisticated breast models containing lumps of varying size, firmness, and depth of placement (e.g., Madden et al., 1978). When skeptics (usually physicians) commented that the silicone models didn't feel exactly like the real thing, Hank liked to quip that an outstanding feature of the flight simulators relied on to train pilots was that they never left the ground! (Among his many skills and interests, Hank was also a licensed pilot.) It was only necessary that the models capture characteristics central to the target discrimination between normal nodularity in breast tissue and tumors—and that they did. Indeed, the models provided for a more complete understanding of the psychophysics of the task; tumor detection proved to be a function of multiple dimensions (i.e., size, firmness, depth within the breast tissue, e.g., Bloom et al., 1982). In doing so, they also provided a platform for identifying the contributions of skilled behavior to tumor detection.

By this point in the research story, Hank had turned attention to a definitive behavioral task analysis. How exactly should the fingers be held and moved to ensure contact with and detection of breast masses? A series of experimental analyses was conducted to isolate each component of breast-examination skill and test which of its variants would yield the greatest number of lump detections in the silicone models and/or the most thorough coverage of the breast tissue. By way of examples, we learned that the flats of the middle three fingers, held together, were the most sensitive and that moving the fingers in small circles with varying degrees of pressure at each location examined were the optimal palpation techniques for finding the simulated masses (e.g., Stephenson et al., 1979). Similarly, a systematic linear pattern of search was superior to alternatives in thoroughness of covering the entire breast area (Saunders et al., 1986). Our confidence in identifying these critical skill components was made possible by Hank's steadfast focus on measurement, which necessarily took some creative turns given the subject matter. To illustrate, exam thoroughness was evaluated by projecting a numbered grid across a supine woman's upper torso via a slide projector on a stand. The projection was then adjusted such that each square of the grid was exactly 1 in, and the number of palpations made in each square during a woman's self-examination was recorded. Areas that went unexamined were captured perfectly, and the resulting record allowed for precise feedback. Observers were calibrated frequently. A central tenet for Hank was that behavior analysts are equipped to take on the most important of social problems because we know how to define, measure, and train behavior, which takes us to the next phase of MammaCare development.

Hank was a teacher, to his very heart—a shaper of behavior par excellence. He had an infectious excitement about the very process of behavior change. This quality has been widely heralded with respect to his classroom teaching, his mentoring, and his influence on Precision Teaching, but it was also the foundation of the breast self-examination teaching system that became MammaCare. As individual skill components were identified in laboratory studies, they were synthesized and incorporated into a targeted, individualized training program for sensitive and thorough examination proficiency. The training package ultimately involved sequential and ongoing practice of each skill component by the learner, first on silicone models and then on herself, with instructions, modeling, prompting, and immediate feedback at each step from lump detection in the silicone models and from a qualified trainer. The training package was developed iteratively and evaluated continuously in terms of culminating final rates of lump detection and thorough self-exams. Even early versions of the training were effective in increasing lump detections (Hall et al., 1980), and the package continued to improve with additional study (e.g., Pennypacker et al., 1982; Pennypacker & Iwata, 1990).

In case it's not obvious by now, Hank was always a big-picture systems-level thinker, as he made evident repeatedly throughout his career. Why write a few papers on measurement and research approaches when you can coauthor a definitive treatment for the field (Johnston et al., 2020)? Why design a few college courses when you can establish an enduring and influential university-level personalized system of programmed instruction? And why be content with improving skills in study participants when you can create a life-saving technology that is available worldwide? The next chapters of the MammaCare story take exactly the latter bent. Armed with a proven teaching system, the question became how best to get it into the hands of women and those who provided for their care while maintaining its quality. Hank was convinced that the private sector could be leveraged to aid in the process of dissemination (e.g., Pennypacker, 1986) and that behavior analysts, in general, took too little advantage of this possibility. Working toward that end became the strategy. Steps included founding a corporation; offering public stock; establishing free-standing and hospital- or medical practice-based training centers with credentialed instructors who had completed a week-long practicum at MammaCare headquarters in Gainesville, Florida; and transferring the teaching program to video-based and then computer-based presentations with automatic feedback that first underwent iterative testing until training outcomes matched those of one-on-one personal training. Today, MammaCare affiliates are located in medical and research centers across the United States, Europe, and the Middle East, and instructor certification programs are ongoing, as are sales of the training package kits. At the same time, research and development has continued to expand the teaching program for instruction of clinical breast exam skills conducted by medical students, residents, nurses, and physicians (e.g., Pilgrim et al., 1993), where its effectiveness continues to be demonstrated to this day.

This brief history of MammaCare has been recounted here because I believe it to be revealing of Hank's legacy—in terms of both its particular accomplishments and in the broader illustration it provides of his manner of leadership within behavior analysis. Hank led by doing, by showing how. And he showed us much. The model he set and the lessons he taught were powerful. They influenced the field as well as the lives of his students and colleagues, and they are as important today as ever—perhaps even more so. Hank entertained no distinction between basic and applied behavior analysis, in spirit or style or conceptual foundations, and his many successes followed from that position. Hank was never afraid to learn by trying—to emit behavior for shaping—and he expected no less of his students and colleagues. If things didn't go quite as planned, Hank's implied counsel was to “rub some dirt on it” and try a better approach. Working with Hank was always an adventure, full of shared wisdoms and witticisms, opportunities, encouragements, and challenges to take the next step. He provided a lasting reminder that we should take our work seriously but not ourselves and that enjoying what we did and who we did it with was paramount. Hank had a wide-ranging intellect and set of interests, a keen curiosity about people and the world, and a mischievous sense of humor. He loved music and travel, a shared laugh or good story and, especially, his wife and family. His was a life well-lived, and our lives are better as well for having had time with him.

Hank was also deeply concerned with the strategies that might ensure our survival as a field. He helped us to appreciate a shared responsibility for service to behavior analysis, and he lived that responsibility in prominent leadership positions until the day of his untimely death. May we all take lessons from Hank's legacy and each do what we can to honor those lessons with our actions. He would wish it no other way.

汉克·彭尼帕克（Hank Pennypacker）对行为科学的发展产生了不可估量的影响。建立和促进行为分析作为一个有凝聚力的领域和专业；佛罗里达大学行为分析项目的基础、健康和活力；以及许多学生和同事的成就，他们的生活和职业受到他的启发，包括我自己。1980 年代初期到中期，当我还是一名研究生时，我有幸成为 Hank 研发团队的一员，当时 MammaCare 的分析和综合工作正在展开。可以说，这一时期值得我们仔细审视汉克的荣誉，因为它代表了他作为学者、教师和个人的遗产的真正缩影。

汉克经常强调最初的实验问题在制定研究计划方面的力量。对于这个特定项目，起源故事在于观察到大多数乳腺肿瘤首先是由女性自己发现的，而不是通过乳房X光检查或超声波等筛查技术发现的。遗憾的是，当肿瘤被发现时，它们通常已经很大，而且如果癌变，已经是晚期了。汉克的问题由此出现：如果手指能够熟练地识别阅读盲文所需的复杂辨别能力，那么在治疗更加有效的情况下，是否可以帮助女性学会检测小肿块？在接下来的 50 多年里，汉克接受了这个问题带来的挑战。为了开始解决这个问题，他首先转向了经典实验心理学的根源。（汉克的博士工作涉及对猴子经典条件反射进行严格控制的实验室研究，众所周知，他将自己描述为“转变的胡利安”。）。在这里，一系列仔细的心理物理学研究详细说明，当手指直接放在目标上时，可以可靠地报告嵌入硅煎饼之间的最微小的滚珠轴承（例如，Adams 等人，  1976），清楚地证明了概念证明。

当然，滚珠轴承不是乳腺肿瘤，并且需要对乳腺组织进行更真实的模拟，因此我们与材料科学家不断合作，开发一种适合辨别测试和训练的介质。汉克经常主张在解决重大问题时与其他学科合作的重要性，这是乳房检查项目的一个最初的例子。由于他对绝对单位的传奇般的坚持，在硅胶中通过迭代过程测量和近似了真实乳腺肿瘤和正常结节的关键尺寸，从而产生了越来越复杂的乳腺模型，其中包含不同大小、硬度和放置深度的肿块（例如，Madden等人，  1978）。当怀疑论者（通常是医生）评论说硅胶模型感觉与真实的东西不完全一样时，汉克喜欢打趣说，用于训练飞行员的飞行模拟器的一个突出特点是它们从未离开地面！（汉克拥有众多技能和兴趣，其中之一是持有执照的飞行员。）模型只需要捕获乳腺组织正常结节和肿瘤之间目标区分的核心特征即可，而且他们确实做到了。事实上，这些模型提供了对任务心理物理学的更完整的理解。肿瘤检测被证明是多个维度的函数（即，乳房组织内的大小、硬度、深度，例如，Bloom等人，  1982）。在此过程中，他们还提供了一个平台来识别熟练行为对肿瘤检测的贡献。

到了研究故事的这一点，汉克已经将注意力转向了明确的行为任务分析。究竟应该如何握持和移动手指才能确保接触和检测乳腺肿块？进行了一系列实验分析，以分离乳房检查技能的每个组成部分，并测试其变体中的哪种变体将在硅胶模型中产生最多数量的肿块检测和/或最彻底地覆盖乳房组织。通过示例，我们了解到，中间三个手指并拢的平面是最敏感的，并且在每个检查位置以不同程度的压力以小圆圈移动手指是寻找模拟肿块的最佳触诊技术（例如，Stephenson 等人，  1979）。同样，系统的线性搜索模式在覆盖整个乳房区域的彻底性方面优于替代方案（Saunders 等，  1986）。汉克对衡量的坚定关注使我们对识别这些关键技能要素充满信心，考虑到主题，这必然需要一些创造性的转变。为了说明这一点，通过通过支架上的幻灯片投影仪在仰卧女性的上躯干上投射编号网格来评估检查的彻底性。然后调整投影，使网格的每个方格正好为 1 英寸，并记录女性自检期间在每个方格中进行的触诊次数。未检查的区域被完美捕获，由此产生的记录可以提供精确的反馈。观察者经常被校准。汉克的核心原则是行为分析师有能力解决最重要的社会问题，因为我们知道如何定义、衡量和训练行为，这将我们带入 MammaCare 发展的下一阶段。

汉克从内心深处是一位老师——一位卓越行为的塑造者。他对行为改变的过程充满了感染力的兴奋。这种品质在他的课堂教学、他的指导以及他对精准教学的影响方面得到了广泛的赞誉，但这也是后来成为 MammaCare 的乳房自我检查教学系统的基础。由于在实验室研究中确定了个人技能组成部分，因此将它们综合起来并纳入有针对性的个性化培训计划中，以提高敏感和彻底的考试能力。培训包最终涉及学习者对每个技能部分的连续和持续的练习，首先是在硅胶模型上，然后是在她自己上，每一步都有说明、建模、提示和即时反馈，从硅胶模型中的肿块检测到合格的合格人员教练。培训包经过迭代开发，并根据肿块检测的最终率和彻底的自检不断进行评估。即使是早期版本的训练也能有效地增加肿块检测（Hall 等人，  1980），并且随着额外的研究，该包继续改进（例如，Pennypacker 等人，  1982；Pennypacker 和 Iwata，  1990）。

也许现在还不是很明显，汉克一直是一个大局观系统级思考者，正如他在整个职业生涯中反复证明的那样。当您可以共同撰写该领域的明确治疗方法时，为什么要写几篇关于测量和研究方法的论文（Johnston 等人，  2020）？当你可以建立持久且有影响力的大学级个性化程序化教学体系时，为什么还要设计一些大学课程呢？当您能够创造出一种全球可用的救生技术时，为什么还要满足于提高研究参与者的技能呢？《MammaCare》故事的后续章节正是采用了后者。有了经过验证的教学系统，问题就变成了如何最好地将其交到妇女和那些为她们提供护理的人手中，同时保持其质量。汉克相信，可以利用私营部门来帮助传播过程（例如，Pennypacker，  1986），而行为分析师总体上很少利用这种可能性。为此而努力成为了战略。步骤包括成立一家公司；发行公开股票；建立独立的医院或医疗实践培训中心，配备在佛罗里达州盖恩斯维尔 MammaCare 总部完成为期一周实习的合格讲师；将教学计划转移到基于视频的演示，然后是基于计算机的演示，并带有自动反馈，首先进行迭代测试，直到培训结果与一对一的个人培训相匹配。如今，MammaCare 附属机构遍布美国、欧洲和中东的医疗和研究中心，讲师认证计划和培训套件的销售也在进行中。与此同时，研究和开发不断扩大由医学生、住院医师、护士和医生进行的临床乳房检查技能教学的教学计划（例如，Pilgrim 等人，  1993 年），其有效性仍然很高。至今已得到证明。

这里重述了 MammaCare 的简史，因为我相信它揭示了汉克的遗产——无论是其特殊成就，还是更广泛地说明了他在行为分析中的领导方式。汉克通过做事、展示如何做来领导。他向我们展示了很多。他树立的榜样和他所传授的教训是有力的。他们影响了这个领域以及他的学生和同事的生活，他们在今天和以往一样重要——也许更是如此。汉克在精神、风格或概念基础上对基本行为分析和应用行为分析没有区别，他的许多成功都源于这一立场。汉克从不害怕通过尝试来学习——通过行为来塑造——他对他的学生和同事也抱有同样的期望。如果事情没有按计划进行，汉克暗示的建议是“擦掉一些污垢”并尝试更好的方法。与汉克一起工作始终是一次冒险，充满了共同的智慧和俏皮话、机遇、鼓励和挑战，以迈出下一步。他持久地提醒我们应该认真对待我们的工作，而不是我们自己，享受我们所做的事情以及与谁一起做是最重要的。汉克有着广泛的才智和兴趣，对人和世界有着强烈的好奇心，还有顽皮的幽默感。他热爱音乐和旅行，喜欢分享欢笑或美好的故事，尤其是他的妻子和家人。他的生活很美好，我们的生活也因为和他在一起而变得更好。

汉克还深切关注可能确保我们作为一个领域生存的策略。他帮助我们认识到为行为分析服务的共同责任，并且他在重要的领导职位上履行了这一责任，直到他英年早逝。愿我们都能从汉克的遗产中汲取教训，并尽我们所能，用我们的行动来纪念这些教训。他不希望有其他方式。