ABSTRACT
According to Pea (2004), the term “scaffolding,” as it applies to educational settings, was first coined in a study conducted by David Wood, Jerome Bruner, and Gail Ross in the mid-1970s. Wood and colleagues wrote: “More often than not, it [problem-solving] involves a kind of ‘scaffolding’ process that enables a child or novice to solve a problem, carry out a task or achieve a goal which would be beyond his unassisted efforts” (p. 90).
Pea, who was a graduate student at the time in Bruner’s Oxford University laboratory, writes that during that period there was great interest in understanding how babies made sense and use of language in structured, but natural one-to-one interactions with a more capable other. In this context, Wood and his colleagues, Bruner and Ross, embarked on a study of children’s problem-solving in one-to-one settings with the underlying notion that the language used by the more capable other would likely matter to the child’s success. Indeed, the title of Wood, Bruner, and Ross’s 1976 paper in which scaffolding was first coined, “The Role of Tutoring in Problem-Solving,” aptly reflects their goal.
It is useful to note that at the time that scaffolding research was getting underway the then current approach to studying development assumed that children developed skills on their own. Wood and his colleagues challenged that prevailing view, describing it this way: “the young learner who went beyond what he or she could do, was viewed as merely imitating or following a model” (Wood, Bruner, & Ross, 1976, p. 90). It may seem obvious to us now that such a view of the learner as unassisted and merely imitating a model must be misguided, however, at the time that Wood and colleagues were studying problem-solving in one-to-one settings, Vygotsky’s work was just being translated and would not be published until 1978, and empirical work on what would be called scaffolding was just beginning (Pea, 2004).
Thus, Wood, Bruner, and Ross’s view that studies of skill development and problem-solving should take in to account the role of language and the tutor was a novel one, and one that made sense in the context of the ongoing lab work studying infants, language, and learning. Indeed, Wood and his colleagues argued, learning occurs in a social context and humans, unlike other species, engage in “intentional tutoring” with children from a very young age. Thus, given our very human capacity for teaching (in other words, helping), it made little sense to Wood and his colleagues to study development as though the learner was alone and unassisted. Instead, they set out to examine the role of the tutor (someone with more knowledge who can help) in helping a learner develop skill in structured one-to-one situations in which the learner is faced with completing a task beyond their unassisted capability.
The approach that Wood and his colleagues took in their 1976 study was an empirical one; that is, they observed a tutor (Gail Ross, the third author) in one-to-one settings, assist 30 three-, four-, and five-year-old children with constructing a toy pyramid built from wooden blocks constructed by the first author, David Wood. Ross was to allow the child to do as much as possible independently, and when help was needed, to first provide verbal instruction and then, if more help was still needed, to intervene more directly, by showing, for example, which block to choose next. The researchers noted all of the tutor’s moves to help each child and the child’s responses to the tutor’s help.
As might be expected in any setting involving humans, Ross did not perfectly follow the protocol of increasing or decreasing support in response to the child’s success or not with assembling the wooden pieces correctly to construct the toy pyramid. Some children received more help when they should have received less: for example, there were instances where Ross showed which block to choose next when instead the child, based on success, should have been given less help, perhaps a verbal instruction such as “Where might that block go?” or “Which block is next?” The expected variation in Ross’s following the protocol with different ages of children, allowed the researchers to study and elucidate effective tutoring. Which children experienced success? Which children had a more difficult time? How did the interactions with Ross differ?
Their observations led Wood, Bruner, and Ross to define the tutor’s role in scaffolding as one that begins with recruiting the learner’s attention and interest in the task, simplifying the task to such a degree that the learner can perceive when an attempt fit or not (being able to recognize when the attempt was wrong or not), maintaining the child’s engagement in the task, marking critical features of the task, minimizing or controlling frustration for the learner, and demonstrating solutions to the task for the learner.
Wood and his colleagues summed up their findings by saying that “Well-executed scaffolding begins by luring the child into actions that produce recognizable-for-him solutions. Once that is achieved, the tutor can interpret discrepancies to the child. Finally, the tutor stands in a confirmatory role until the tutee is checked out to fly on his own” (p. 96).
