学习的本质：认知科学视角下的知识建构

学习的本质：认知科学视角下的知识建构

学习是人类认知活动的基础，也是教育科学研究的核心议题。然而，当我们追问"学习究竟是什么"这一根本问题时，却发现这个看似简单的问题远未得到令人满意的回答。在日常语境中，"学习"一词被赋予了过于宽泛的含义——从记忆单词、练习钢琴到理解相对论，似乎都可以归入"学习"的范畴。这种概念的模糊性不仅阻碍了学习科学的深入发展，也给教育实践带来了诸多困惑。本章将从认知科学的角度出发，系统地重新审视学习的本质，试图为这一古老而常新的问题提供更为精确的理论框架。

The nature of learning has been a central concern in cognitive science, yet the question "what is learning, exactly?" remains surprisingly elusive. In everyday discourse, the term "learning" encompasses an enormous range of phenomena—from memorizing vocabulary and practicing piano to comprehending relativity. This conceptual vagueness not only impedes scientific progress in understanding learning but also creates substantial challenges for educational practice. This chapter approaches this fundamental question from the perspective of cognitive science, seeking to establish a more rigorous theoretical framework for understanding the essence of learning.

从行为主义到认知革命：学习研究的历史演进

行为主义范式的兴起与核心假设

理解当代学习理论，必须首先回溯其历史脉络。现代学习研究的科学化进程始于20世纪初的行为主义（Behaviorism）革命。在此之前，学习研究主要由哲学思辨主导，缺乏系统的实证方法。行为主义的创始人约翰·华生（John B. Watson）在1913年发表的《行为主义者眼中的心理学》一文中，明确提出心理学应当成为一门纯粹客观的实验性自然科学，研究的对象应该是可观察的外显行为，而非不可捉摸的内心意识。这一宣言标志着行为主义作为一个独立研究范式的正式确立。

The scientific study of learning began with the behaviorist revolution in the early twentieth century. Before this paradigm emerged, learning research was dominated by philosophical speculation lacking systematic empirical methods. John B. Watson's 1913 article "Psychology as the Behaviorist Views It" declared that psychology should become a purely objective experimental natural science, focusing exclusively on observable behavior rather than unobservable consciousness. This manifesto established behaviorism as an independent research paradigm.

行为主义的兴起有其深刻的哲学和方法论背景。在哲学层面，实证主义（Positivism）思想强调科学知识必须建立在可观察、可验证的经验基础之上，将一切不可直接观察的实体——如心灵、意识、意图——视为形而上学的虚构予以排斥。在方法论层面，行为主义坚持方法论还原论（Methodological Reductionism），认为复杂的心理现象最终可以还原为简单的刺激-反应联结，通过研究后者的规律即可理解前者。这些哲学预设决定了行为主义对学习现象的基本理解方式。

The rise of behaviorism was grounded in deep philosophical and methodological commitments. Philosophically, positivism emphasized that scientific knowledge must rest on observable and verifiable empirical foundations, rejecting as metaphysical fictions any entities—such as mind, consciousness, or intention—that could not be directly observed. Methodologically, behaviorism embraced methodological reductionism, holding that complex psychological phenomena could ultimately be reduced to simple stimulus-response associations, and that understanding the latter would illuminate the former.

在行为主义框架中，学习被定义为有机体在经验基础上发生的相对持久的行为变化。华生的追随者爱德华·桑代克（Edward L. Thorndike）通过大量动物实验提出了著名的效果律（Law of Effect）：动物在某一情境中做出的行为如果带来了满意的结果，该行为在类似情境中再次出现的概率就会增加；反之，如果行为带来的是不满意的结果，该行为出现的概率就会降低。效果律的发现为理解学习提供了一个简洁而有力的解释框架。

Within the behaviorist framework, learning was defined as a relatively permanent change in behavior resulting from experience. Watson's follower Edward L. Thorndike proposed the famous Law of Effect through extensive animal experiments: behaviors followed by satisfying consequences in a given situation become more likely to recur in similar situations, while behaviors followed by unsatisfying consequences become less likely. This principle provided a simple yet powerful explanatory framework for understanding learning.

另一位对行为主义发展做出关键贡献的是B.F.斯金纳（B.F. Skinner）。斯金纳在桑代克效果律的基础上，发展出了一套更为精致的学习理论——操作性条件作用（Operant Conditioning）。与桑代克主要关注的工具性学习不同，斯金纳将研究重心放在有机体自发行为与环境结果之间的功能关系上。他认为，有机体的许多行为并非由特定刺激所诱发，而是在环境中自发产生的"操作"（operant）。这些操作的命运——是受到强化（reinforcement）还是惩罚（punishment）——决定了它们未来出现的频率。斯金纳进一步区分了正强化（positive reinforcement）和负强化（negative reinforcement）、积极惩罚（positive punishment）和消极惩罚（negative punishment）四种基本操作，并系统地研究了各种强化程序（schedules of reinforcement）对学习效果的影响。

Another pivotal contributor to behaviorism was B.F. Skinner, who developed a more sophisticated learning theory called operant conditioning building on Thorndike's Law of Effect. Unlike Thorndike's focus on instrumental learning, Skinner concentrated on the functional relationships between organisms' spontaneous behaviors and environmental consequences. He argued that many behaviors are not elicited by specific stimuli but rather emerge spontaneously as "operants" in the environment. The fate of these operants—whether reinforced or punished—determines their future frequency. Skinner further distinguished among four basic operations: positive reinforcement, negative reinforcement, positive punishment, and negative punishment, systematically investigating how various reinforcement schedules affect learning outcomes.

行为主义的局限性与理论困境

行为主义在20世纪上半叶取得了压倒性的优势，几乎垄断了学习研究的学术话语权。然而，随着研究的深入，行为主义范式逐渐暴露出其固有的理论局限。这些局限不仅动摇了行为主义的统治地位，最终还催生了学习研究的认知革命（Cognitive Revolution）。

Despite its dominance in the first half of the twentieth century, the behaviorist paradigm eventually revealed fundamental limitations. These constraints not only undermined behaviorism's preeminence but ultimately catalyzed the cognitive revolution in learning research.

首先，行为主义将一切心理现象还原为外显行为的做法，导致了学习概念的过度简化。人类的学习过程涉及复杂的认知加工——注意、编码、存储、提取、迁移——这些过程虽然无法直接观察，却在学习活动中发挥着不可替代的作用。行为主义将所有这些中间变量斥为"黑箱"而不予研究，实际上放弃了对学习机制本身的解释。例如，一个学生解数学题时，是通过什么认知过程找到解题方法的？不同学生为何在同一教学条件下学习效果差异巨大？行为主义的刺激-反应框架根本无法回答这类问题。

First, behaviorism's reduction of all psychological phenomena to overt behavior led to an oversimplified conception of learning. Human learning involves complex cognitive processes—attention, encoding, storage, retrieval, transfer—that, while unobservable, play indispensable roles in learning activities. By dismissing these intermediate variables as a "black box" unworthy of investigation, behaviorism effectively abandoned any attempt to explain the mechanisms underlying learning. For instance, what cognitive processes does a student employ when solving a math problem? Why do different students show vastly different learning outcomes under identical instructional conditions? The stimulus-response framework of behaviorism is fundamentally incapable of addressing such questions.

其次，行为主义的理论预测力极为有限。效果律和操作性条件作用原理虽然能够解释某些简单的学习现象，但在面对复杂的人类学习任务时显得苍白无力。例如，语言习得（language acquisition）是人类最显著的学习能力之一，但行为主义将儿童习得语言的过程理解为对语音刺激的模仿和强化，这一解释无法说明儿童为何能够在有限的经验基础上创造出无限多的新句子，也无法解释语言习得的敏感期现象。

Second, behaviorism's theoretical predictive power proved severely limited. While the Law of Effect and operant conditioning principles could explain certain simple learning phenomena, they proved inadequate when confronting complex human learning tasks. Language acquisition represents one of humans' most remarkable learning capabilities, yet behaviorism's explanation of children learning language through imitation and reinforcement of speech stimuli fails to account for children's ability to generate infinitely many novel sentences from limited experience, nor can it explain critical periods in language development.

第三，行为主义在方法论上的严格限制阻碍了学习研究与其他认知科学领域的整合。对内隐认知加工的排斥使得行为主义无法与认知心理学、神经科学、计算科学等邻近学科进行有效对话，形成了一个自我封闭的理论孤岛。当20世纪中叶计算机科学的兴起为认知研究提供了新的隐喻和工具时，行为主义却因为坚持其反心理主义的方法论立场而错失了参与认知科学革命的良机。

Third, behaviorism's methodological restrictions prevented integration with other cognitive science disciplines. The rejection of implicit cognitive processing isolated behaviorism from productive dialogue with cognitive psychology, neuroscience, computational science, and neighboring fields, creating a self-contained theoretical enclave. When the rise of computer science in the mid-twentieth century provided new metaphors and tools for studying cognition, behaviorism—constrained by its anti-mentalist methodological commitments—missed the opportunity to participate in the cognitive science revolution.

认知革命的兴起与范式转变

20世纪50年代末至60年代初，一股新的学术潮流开始在心理学领域涌动，最终汇聚成为一场深刻的认知革命。这场革命的到来既有内外部学科发展的推动，也源于对行为主义局限性的深刻反思。

In the late 1950s and early 1960s, a new intellectual current began to emerge in psychology, eventually coalescing into a profound cognitive revolution. This transformation was driven by both internal and external disciplinary developments, as well as by growing recognition of behaviorism's limitations.

认知革命