1.3.1 First Principle: Mediation

Mediation: a “transition from direct, innate, natural forms and methods of behavior to mediated, artificial mental functions that develop in the process of cultural development” [italics in original] (Reference VygotskyVygotsky, 1998, p. 168). This principle reflects “the historical development of human behavior” (p. 168) as cultures create different kinds of physical and symbolic tools to help them modify nature to survive and to improve their living conditions. For example, if we need to dig a hole to plant a tree, we can try using our hands as an animal might. However, it is much easier and more efficient to use a shovel – a cultural invention that enhances our ability to dig holes. If we need to plant many trees in a short period of time, we can use specially designed power machines – a more advanced version of hands/arms and shovels. Similarly, if I want someone, such as a child, to move from one location to another, I can do so by physically pushing or pulling the person, or I can use another human creation – language – to achieve my goal symbolically. Both physical and symbolic tools, such as shovels and language are said to mediate human relationships to material reality and to each other. In other words, culturally created tools come between myself and reality (physical or social) and in so doing, change my relationship to this reality. For a fuller discussion of mediational tools, see Reference KozulinKozulin (2024).

1.3.2 Second Principle: Sociogenesis

Sociogenesis: “the relation between higher mental functions was at one time a concrete relation between people; collective social forms of behavior in the process of development become a method of individual adaptations and forms of behavior and thinking of the personality” [italics in original] (Reference VygotskyVygotsky, 1998, p. 168). This principle captures the fact that every higher mental process appears twice in development, first interpersonally, between people, and then intrapersonally, within each individual. Thus, functions such as attention, perception, memory, imagination, emotions, logical thinking are in the beginning of our life mediated through our social relationships with other people. In early life, the other people are usually parents and older siblings. Later, as we grow, other individuals take on this role, including teachers, friends, co-workers, and so forth. How other people behave toward us and eventually how we behave toward them interpersonally affects how we relate to ourselves psychologically, or intrapersonally. This process takes place primarily through linguistic communication. A simple example having to do with perception should make things clearer. In early childhood, before we learn language, our natural perceptual instinct enables us to perceive objects and movements around us. However, prelinguistic children do not have any idea what they are perceiving until another person, such as a caregiver, tells them. Caregivers usually do not do this with a formal plan in mind, in the way a teacher might plan a lesson; nevertheless, it is through speaking with children that their natural ability to perceive is reshaped as a culturally constructed way to perceive. A child might see an object such as a drinking glass sitting on a table. The child has no way of knowing that the glass and the table are separate entities and might easily conclude that the glass is part of the table, until someone lifts and refers to the object as “glass” and also hears the other person refer to the object on which it was resting as “table.” Over time, the child is likely to see “glass” in a variety of different contexts and will then come to understand that it is not a feature of table but is in fact a separate object, with a specific function.

As our language system develops, more and more parts of reality become visible to us as they are mediated through the meanings of our language, which is why words are important units of analysis for Vygotsky. Not only are we then able to talk about these parts, but we eventually are able to think about them as well. Thus, what was in the beginning perception mediated by others, interpersonally, becomes perception that is self-mediated, intrapersonally, but relying on the same symbolic meanings that others presented to us. Reference Vocate and VocateVocate (1994) described the shift from social communication between people to psychological communication within an individual as movement from “I ~ You” dialogues to “I ~ Me” dialogues.

Not only do we learn how to think about reality through mediation provided by others, we also learn how to feel as a consequence of “I ~ You” dialogues. We are born with a set of emotions that have developed over the course of human evolution. These include among others, distress, exhibited through crying as when an infant has an “urgent need” for food (Reference HolodynskiHolodynski, 2013, p. 24) and endogenous pleasure, displayed through a smile and the relaxing of tension upon recognizing a caregiver’s face (p. 25). During the enculturation process these natural emotions, also mediated by others, develop into more complex emotions as they are brought into language. Distress, for example, in Western cultures, morphs into frustration, anger, defiance, sorrow, and sadness (Reference HolodynskiHolodynski, 2013, p. 26), while endogenous pleasure develops into pleasure, joy, affection, and amusement (p. 26). Just as with perception, once emotions are brought into language and become semantic emotions and not just biological feelings, we are able to understand what it is we feel as defined by our culture. Moreover, we can use language to talk about our emotions with others even if we do not experience them at the time. Indeed, before his untimely death, Vygotsky proposed a new unit of analysis for understanding consciousness – perezhivanie – the unity of intellect/cognition and emotion (see Reference Vygotsky, van der Veer and ValsinerVygotsky, 1994).

1.3.3 Third Principle: Internalization

Internalization: the “transition of a function [is] from outside inward” [italics in original] (Reference VygotskyVygotsky, 1998, p. 170). The point of this principle, especially as it relates to principle 2, is that our psychological makeup is shaped by our social experiences. Thus, the connection between our psychology and our social world is necessary and inseparable. Humans as such are always social, even when alone because we always carry others with us in our I ~ Me dialogue what was originally I ~ You dialogue. Reference VygotskyVygotsky (1997b, p. 170) stated this connection cogently and succinctly: “through others, we become ourselves.”

The process alluded to in the third principle is described as internalization, which entails the use of language (i.e., speaking) to master, or regulate, our own mental and physical behavior. In so doing we gain freedom from coercion by the here and now of the immediate environment. Children are not able to understand and talk about the past and the future or about things not in their immediate surroundings until they have developed a sufficiently sophisticated language system that includes the means of talking and thinking about past and future objects and events – that is, things that are extra-contextual. This capacity is crucial in the development of free will, which Reference VygotskyVygotsky (1997b, p. 171) equates with verbal behavior: “without speech, there is no will.”

An important aspect of internalization is that it not only encompasses cognitive development but also entails the formation of culturally appropriate emotions that emerge from precursor natural emotions (i.e., distress, disgust, interest, startle, and pleasure) as mentioned in Section 1.3.2 (see Reference HolodynskiHolodynski, 2013, pp. 24–25). As children are enculturated along with their cognitive development, their natural emotional instincts are linguistically restructured into a semanticized emotional system as they engage with members of their community. Thus, not only are they able to perceive natural objects as different as palms, pines, oak, and maples as belonging to the same superordinate category “tree,” they are also able to perceive, think about, and linguistically express specific bodily reactions and feelings such as love, hate, anger, fear, shame, joy, and so forth, even when they are not actually experiencing these in real time. Concepts such as “tree,” as argued by Reference DanzigerDanziger (1997), are not natural kinds, but are human kinds created throughout cultural history as communities interacted with nature. In other words, what is expressed through the English word tree is not a specific concrete object that would exist in nature in the absence of humans. What exist in nature are concrete objects referenced in English as palms, maples, pines, oaks, cypresses. Moreover, not all cultures have found it necessary to group the array of natural woody objects into a single category. Some indigenous communities in Australia do not have a word equivalent to English tree, but instead they have individual words for each of the wide variety of eucalyptuses growing in their environment. Similarly, semanticized emotions created from the precursor emotions are not universal across human cultures. For example, Reference RatnerRatner (1990, p. 78) points out that according to some anthropologists, indigenous North American Arctic communities lack the concept of “anger” “because they do not blame individuals for their actions,” even though they may feel annoyed at a particular act perpetrated by someone else.

1.3.4 Fourth Principle: Developmental Stages

Developmental Stages: as every higher mental function is internalized it passes through four developmental stages (Reference VygotskyVygotsky, 1997b, p. 103). The first is the stage where our natural innately specified instincts predominate. The second is the stage of external mediation of behavior by other people. It initiates the organization and subordination of the natural processes through social “I ~ You” communication. The third stage occurs when the individual begins to direct the symbolic means used by others toward the self in “I ~ Me” communication in order to mediate, or regulate, our own mental and physical behavior. This stage is associated with egocentric or private speech (Reference VygotskyVygotsky,1987; Reference FlavellFlavell, 1966) – speech that at first appears to be social in form but is psychological in function. The fourth stage appears when private speech is completely internalized and no longer overtly expressed. At this point, it becomes inner speech, which loses all the formal features of external speech but retains its meaning as it mediates our mental activity.

To give an example of private speech, consider someone trying to complete a jigsaw puzzle. As the person works on the puzzle, it would not be unusual for her to produce such utterances as “Now, the red one,” or “Wait, wrong,” “Blue.” As social speech, these utterances are difficult to understand, because they are not intended for an interlocutor. In other words, they are part of an “I ~ Me” dialogue in which the speaker directs (or mediates) herself through the difficulty of deciding which piece needs to be placed in a particular position. The speaker knows that “red” and “blue” refer to puzzle pieces of a particular color and she knows that “wrong” means something like “I made the wrong selection and I need to make a different one,” but because she is engaged in a private dialogue it is not necessary to form a full utterance, as might happen in social speech. To be sure, social dialogues, given presuppositions that can be drawn on the basis of context and previous utterances, can take on an abbreviated appearance, also, as when two people are looking at a painting in a museum and one says to the other “beautiful.” In the puzzle example, however, there is no interlocutor other than the self. Moreover, the motivation behind the puzzler’s utterances is the problem of determining which is the appropriate piece to place in the puzzle at this point in the process. If there were no such problem, it is unlikely that she would produce such utterances and the task would be planned internally without any need to externalize the thinking process. The reason that the process is externalized is that when we encounter psychological difficulties, we tend to “reaccess” (Reference Frawley and LantolfFrawley & Lantolf, 1985) earlier stages of development, as stated in the four principles. The following quote from Reference VygotskyVygotsky (1997c, p. 106) captures the essence of the theory:

1.4.1 The Genetic Method and Education

Researchers in the hard and social sciences generally accept the segregation of theory/research from its practical applications whereby theory development and associated research are carried out independently and without regard for possible implications for making practical improvements in the life of a community. Accordingly, the goal of theory/research is to explain phenomena in any domain of reality that is hidden from observation. If the findings turn out to be practically relevant, so much the better, but if not, it does not matter. Vygotsky argued that there is no reason to segregate theory/research from practice in the social sciences in general, and in psychology in particular. For him, the ultimate arbiter of theory is not experimental research (although as we have discussed, it is important), it is practice, as he forcefully stated in the following quote:

In the foregoing remarks, Vygotsky proposes a radical shift in how psychology should pursue research – one in which it can no longer segregate practical application of research from research itself. In other words, the application is not what we do after we have conducted research independent of practice; rather, practice itself is research – research that assesses the validity of the theory. Recall Luria’s use of external mediation with the patient experiencing problems producing connected speech. Was he engaged in research or in clinical practice? Luria relied on the theoretical principle that this form of mediation is essential in the developmental process. If the patient had not responded favorably after writing down the unordered fragments relevant to the theme, it would have represented a significant challenge to the principle.

A practical domain that Vygotsky was intently interested in was education, which he considered to be not merely an activity concerned with knowledge acquisition, but in fact a crucial site for development of the person. As such, the principles of the theory are expected to operate as much in an educational environment as they do in everyday life but in a way that is distinct from everyday life. In other words, education involves a particular use of the instrumental method that intentionally and systematically brings to bear a form of mediation based on the conceptual knowledge that research in all domains, including the arts and humanities as well as the soft and hard sciences, has produced over the course of human history. This form of mediation Vygotsky called “scientific” or “theoretical” which he contrasts with “everyday” or “spontaneous” knowledge (Reference VygotskyVygotsky, 1987). Scientific knowledge reveals aspects of reality hidden from direct observation through our senses. We have already mentioned several examples of scientific knowledge derived from research contrasted with everyday knowledge based on appearances (e.g., whales and fish; fruits and vegetables). Scientific knowledge not only allows us access to the hidden features of reality, it also provides vicarious access to human history (e.g., the French Revolution, feudalism, Roman empire, etc.) and to locations that most of us are unlikely to experience firsthand (e.g., Antarctica, deep sea, surface of the moon). Education accomplishes its task largely through systematically created, organized, and sequenced linguistic signs, “designed by an external agent,” such as teachers, textbooks, syllabus, curriculum, and so forth (Reference Wertsch, Daniels, Cole and WertschWertsch, 2007, p. 185). Given its systematic, intentional, and explicit, focus, Reference VygotskyVygotsky (1997b, p. 88) characterizes education as the “artificial development of the child,” which “restructures all functions of behavior in a most essential manner.” As students move through school, they not only acquire knowledge, which for Vygotsky is not the most important aspect of the educational process, but they enhance the “degree of mastery” (p. 88) they exhibit over their own behavior, which he regarded as the raison d’être of schooling.

1.4.2 Language Education

Over the years we have witnessed a swing of the pendulum in discussions of education, including language education, from so-called teacher-centered to learner-centered approaches. Currently, the pendulum seems to be stuck in the learner-centered position, where students are assumed to be, or at least encouraged to become, the agents of their own autonomous learning, usually in classrooms where explicit instruction, if it occurs at all, is relegated to a marginal role with implicit learning in various formats, such as VanPatten’s model of processing instruction (Reference VanPatten and SmithVanPatten & Smith, 2022), or the Dynamic Usage-Based model proposed by Reference Verspoor, Schmid and McManusVerspoor and Schmid (2024).

Education, including language education, has been influenced directly or indirectly by the writings of Swiss psychologist Jean Piaget, who proposed that development was governed by natural spontaneous processes which unfold in a specific sequence from simple, concrete, and empirical cognition to complex, abstract, and rational thinking (Reference WadsworthWadsworth, 1984). Schooling then should take advantage of the assumed developmental sequence when it comes to organizing and implementing a curriculum in any subject matter. If instruction is not subordinated to natural processes, it will be ineffective and might even do harm to students. Consequently, students must be allowed to discover things on their own when they are ready and no amount of teaching can make a difference other than in trivial ways (Reference EganEgan, 2002, p. 104). According to Reference EganEgan (2002, p. 106), the problem with Piaget’s model is that it focuses on responding to what students can do at any point in the educative process rather than on challenging them and helping them to pursue what they cannot yet do. A study by Reference Kirschner, Sweller and ClarkKirschner, Sweller, and Clark (2006) analyzing research on different versions of discovery- and inquiry-based instruction (a legacy of the natural-child model of education) concluded that there is no significant evidence to support the approach, and that it may even have a negative impact “when students acquire misconceptions or incomplete or disorganized knowledge” (p. 84).

In language instruction, Piaget’s ideas have at least indirectly influenced assumptions in the L2 field regarding so-called natural orders (Reference KrashenKrashen, 1982), in which certain features of an L2 are predicted to be acquired before other features (e.g., English irregular past tense forms acquired before regular past tense forms) and developmental sequences (Reference PienemannPienemann, 1998), in which the acquisition of a specific structure such as negation or question formation in English unfolds in a specific sequence. Thus, in acquiring English negation learners are predicted to follow a predetermined sequence in which they first mark negation with “no/not” placed at the front of a sentence, as in “No John eat dinner”; later they insert “no” or “not” in front of the main verb, as in “John no/not eat dinner” and both of these are prerequisite stages for the auxiliary “do” to appear, as in “John does not eat dinner.”

On analogy with child language acquisition, the natural order and acquisition sequence hypotheses both presuppose that learner-internal mechanisms are responsible for guiding the L2 acquisition process whether it occurs in a natural immersion setting or in a formal classroom context (Reference LongLong, 1997). Bearing this in mind, as with Piagetian educational theory, the claim is that instruction can only be effective if it is sensitive to those mechanisms that developmentally prepare learners to acquire the next feature in a natural order or the next stage in an acquisition sequence. Carrying through on this view, implicit instruction that provides a robust quantity of comprehensible exposure to a language should be more conducive to successful acquisition than is explicit instruction focused on specific features of the language. In other words, focus is on the learner and the learning process with the teacher having responsibility for providing learners with sufficient exposure to samples of the relevant features in natural comprehensible contexts that allow the internal mechanisms to operate.

The goal of implicit instruction, however, is based on two problematic assumptions: that so-called natural learning – how we learn outside of the tutored setting – is superior to well-organized intentional instruction (see Reference EganEgan, 2002 for a detailed critique of natural learning in schools), and that the adult mind is not qualitatively different from the child mind. While adults might be capable of acquiring a new language through implicit processing in everyday immersion settings and through implicit forms of classroom instruction, the process is time consuming and often results in incomplete acquisition (Reference ParadisParadis, 2009, p. 118). There are a number of reasons that might account for incomplete acquisition: some features of the new language may not occur with sufficient frequency to allow learners to inductively figure out their meaning and how they function (e.g., some word-order options in Chinese); some features may be too subtle to detect in the high-speed world of day-to-day communication, even if they are frequently used (e.g., English articles); “attentional biases” emanating from a learner’s L1 that can hinder the ability to notice L2 features (Reference Ellis, Wulff, VanPatten, Keating and WulffEllis & Wulff, 2020); evidence from cognitive neuroscience research that the system responsible for implicit learning in childhood, procedural memory, declines as we age making it more likely that adults will increasingly rely on the system responsible for explicit learning, declarative memory (Reference ParadisParadis, 2009, p. 118). We return to this topic in our discussion of teacher education in Section 3.

SCT-informed language instruction takes a very different perspective in proposing that optimally effective pedagogy depends on well-organized systematic instruction that presents learners with conceptual knowledge of the language, especially of complex and subtle features that are difficult to appropriate from immersion context outside or inside classrooms. The knowledge must be made functional for communicative purposes, which requires not only practice opportunities but also effective mediation that guides learners as they undertake to gain control over and use the new language to create meaning.

2.2.1 Phase 1: Pre-understanding

The Orienting Basis of Action (OBA) is the phase in which an action plan is formulated based on the current knowledge of students. The plan orients an action toward a specific goal and is “the most important aspect of the psychological mechanism of an action” (Reference Gal’perin, Cole and MaltzmanGal’perin, 1969, p. 251). Most of the time we act intentionally based on a plan that we formulate; we execute the plan and we monitor it to determine if the plan is adequately carried out or if it needs to be adjusted in some way. Keep in mind that a plan is a kind of action carried out mentally but not in reality. We can think of a plan as an instruction to ourselves to do something in a specific way. However, we can never know for sure how good the plan is until it is executed. Thus, there is a connection between planning, the focus of phases 1–3 and performing, phase 5.

Phase 1 is important because it informs the teacher and makes students aware of what they know or think they know about a language feature and how they use it to plan and carry out communicative action. Understanding what can be communicatively achieved (or not) through use of a feature of a new language is essential to the planning process. To be communicatively effective a language user must understand the expressive range that a language feature has the capacity to express. If learners’ knowledge is limited, incomplete, or erroneous, their ability to effectively plan and execute communicative actions will be impeded. Problems with their knowledge can result from previous instruction where structure-based rules of thumb are presented. It may also occur through immersion experiences in which a feature is neither sufficiently frequent nor robustly salient, or if the feature was misinterpreted by a learner. The following is an example of a rule of thumb: use Spanish preterit whenever a past temporal adverb is used, as in Ayer fui (preterit) al cine con mis amigos “Yesterday, I went to the movies with my friends.” While the rule is appropriate for the particular context, it does not reflect the option of using imperfect aspect with the same temporal adverb, as in Ayer el presidente hablaba (imperfect) con los líderes europeos “Yesterday, the president spoke/was speaking with the European leaders.”

2.2.2 Phase 2: Concept Presentation

Coherent explanation of a concept so that learners can understand it and when possible, compare it to their pre-understanding revealed in phase 1. The concept must focus on the meaning of each feature rather than its structure. Of course, the structure matters, but C-BLI finds its inspiration in research that has been carried out in meaning-based, rather than structure-based, theories of language – in particular, Cognitive and Systemic Functional Linguistics (CL or SFL). While we find inspiration in these theories, for pedagogical purposes, it is at times useful to modify details of the theoretical explanations to ensure learner understanding. To appreciate the relevance of meaning-based language analysis, we will offer an example borrowed from Reference LangackerLangacker (2008).

According to Reference LangackerLangacker (2008, p. 43), meaning is comprised of content and the specific way a user construes, or chooses to view or perceive, the content. A glass containing water is observed by different speakers of English. One speaker may be motivated to attend to, or construe, the object as “the glass of water.” Another speaker may be motivated to construe the content of the glass and decide to express this as “the water in the glass.” Two other speakers might attend to the quantity of water in the glass. However, in this case, there might be a difference in their construal of the quantity. One might see the glass as “half empty” and the other as “half full” (Reference LangackerLangacker, 2008, p. 44). Speakers express, or profile, their construals through different linguistic means. This informs a listener of the particular aspect of the object that a speaker has in focus, which in turn influences how a listener might construe the object. The teacher’s task is to impart to students the ways in which a language enables speakers to profile their particular construal of objects, events, and states.

We can also construe events in a variety of ways. For instance, consider a simple event such as the following: person(X)/action (left)/object (her lunch)/location (at home)/frequency (for the third time this week). A speaker might construe this event in a variety of ways, such as a more or less neutral way without highlighting any particular component of the event: “Sally left her lunch at home for the third time this week.” However, a speaker might construe the frequency component of the event as salient, in which case she might profile that component by positioning its representation at the front of the utterance, as in “For the third time this week, Sally left her lunch at home.” Another speaker might construe the object and profile it by positioning its representation at the front of the utterance as in “Her lunch, Sally left it at home for the third time this week.” Notice that in English, when a speaker profiles an object, it often entails use of what is called a resumptive pronoun (e.g., it). This can occur in response to a question such as “What happened to Sally’s lunch?” It would, however, be an odd response to the question “What happened?” In this case an appropriate response would be “Sally left her lunch at home for the third time this week.” Finally, a speaker might choose to construe Sally as the salient component of the event, in which case she would be profiled as in “Sally, she left her lunch at home for the third time this week.” This utterance would not be an appropriate response to the question “What happened?” Also notice use of the resumptive pronoun “she.” The meaning of the utterance is quite different from an utterance that does not profile Sally (example borrowed from Reference LangackerLangacker, 2008).

The process that profiles a constituent of an utterance, as in the examples in the preceding paragraph, is referred to as “topicalization.” It has the effect of indicating that the profiled constituent is the topic of the utterance – what the utterance is about. In the preceding examples, we saw that it could be how frequently Sally left her lunch at home; or the object that Sally left at home, or even who left her lunch at home. Languages can topicalize constituents in a variety of ways, but word order is a frequent way of doing it. As we will see when we discuss phase 3, a language such as Chinese profiles a topic in some ways similar to, but in some ways different from, English.

To summarize: phase 2 is the point at which a language concept is systematically presented to students. The teacher needs to be sure that the explanation is understood by students. However, it is important that the goal is not for students to memorize the explanation but to understand it. In fact, in phase 3 the concept is presented to students in a nonverbal visual way, which, if prepared appropriately, will avoid the tendency to memorize without understanding.

2.2.3 Phase 3: Materialization

In this phase the teacher presents students with a visual representation of the concept explained in phase 2. This representation is known as a Schema for the Orienting Basis of Action (SCOBA), or Schema for the OBA. Recall that orientation is central to the planning process. It is the point at which a person surveys a situation, determines, if and how to (re)act, assesses the resources available to (re)act and then develops a plan of action. SCOBAs represent the linguistic resources (conceptual knowledge) available for communicative action. They represent the resources in a holistic external form. Recall that development is understood as a process whose source is found in socially organized relationships and/or culturally created artifacts. SCOBAs are intended as a way of complying with this principle. Because they are visual, as well as tactile (see Figure 1), SCOBAs are crucially holistic, something that is difficult to achieve in verbal explanations. If prepared properly, SCOBAs avoid the tendency of students to memorize, often without understanding, as often happens with verbally presented definitions and explanations.

Figure 1 Tactile SCOBA for Chinese topicalization

In a study on teaching Chinese topicalization reported in Reference Zhang and LantolfZhang and Lantolf (2015), the instructor used two different SCOBAs to visualize the concept for his students. One was a power point animation depicting someone who ate a bowl of rice at home at a specific time of day. The animation showed how to profile the adverbs of time and place as well as the object (bowl of rice) by placing them in the appropriate topic position. In addition, the instructor also used a tactile SCOBA, illustrated in Figure 1, that enabled the students to manipulate the topicalization process with their hands. Depicted in the figure are Cuisenaire rods, often used in teaching math and associated with the Silent Way methodology (Reference GattegnoGattegno, 1972).

The students manipulated the rods in order to demonstrate the various options for profiling different constituents in Chinese utterances. The verb in Chinese cannot occupy any other position, which is indicated by the size of the rod – the verb cannot move. Nevertheless, it is possible to profile a verb in Chinese utterances, but how this is done is beyond the scope of our present discussion.

The study challenged the teachability hypothesis, which claims that learners follow a specific sequence when acquiring certain features of a new language, such as the formation of English negative constructions (Reference PienemannPienemann, 1989). For Chinese, the hypothesis asserts that topicalization occurs in three stages. The first is the normal canonical S (Adv) (Adv) V O order, “The man at 2 pm at home ate a bowl of rice.” The second is the ability to topicalize adverbs, Adv S (Adv) V O, as shown in Figure 1: “At 2 pm the man at home ate a bowl of rice,” or “At home the man at 2 pm ate a bowl of rice.” The third involves positioning the object in topic position O S (Adv) (Adv) V, “A bowl of rice, the man at 2 pm at home ate.” According to the hypothesis, acquisition proceeds through three stages: 1>2>3. Stages cannot be skipped; thus, even under instruction 1>3>2 is not possible and stages 2 and 3 cannot be acquired at the same time. Nonetheless, Reference Zhang and LantolfZhang and Lantolf (2015) reported that through C-BLI students could develop the capacity to spontaneously use stage 3 before stage 2 and that it was even possible to develop the ability to use both stages simultaneously (see also Reference ZhangZhang, 2020). The point is not that C-BLI is some kind of magic bullet but that mechanisms of development are not contained within the individual; rather, they are situated between the individual and the social environment, which if appropriately organized can affect developmental processes.

A particularly interesting outcome of the Chinese study is the performance of one of the students, who was assessed as having a low level of working memory (WM), the mental system that “keeps things in mind while performing complex tasks such as reasoning, comprehension and learning” (Reference BaddeleyBaddeley, 2010, p. 136). On narrative and interview tasks that assessed the learners’ ability to topicalize appropriately in spontaneous speech, the student performed as well as students with normal and high WM. When this student spoke in Chinese, however, she moved her hands as if she were manipulating the Cuisenaire rods, even though the rods were not physically present (Reference Lantolf, Zhang, Loewen and SatoLantolf & Zhang, 2017). This is not a surprising outcome, given the mediational principles considered in Section 1. Indeed, as Reference EganEgan (2002) points out, while our hands are not part of our brain, they indeed can function as part of our mind. The study provided additional evidence that external forms of mediation, if appropriately organized, can enhance the developmental process.

A more recent study (Reference KisslingKissling, 2023) on C-BLI and Spanish verbal aspect provides further evidence that this approach to explicit instruction can significantly alter the anticipated course of L2 development that is assumed to depend on learner-internal mechanisms. Reference Andersen, Huebner and FergusonAndersen (1991) proposed the Aspect Hypothesis (AH), which claims that all L2 learners will adhere to a specific predictable sequence when acquiring languages that morphologically mark aspectual differences. Spanish is one of those languages as it formally marks the distinction between perfective and imperfective aspect. According to the AH, learners will first mark perfective and imperfective aspect that matches the lexical aspect of a verb. Without getting into details (see Kissling for fuller explanation), verbs such as jump, throw, and hit indicate actions that are inherently bounded, in which the action occurs instantaneously (see in Section 2.3.2 for a SCOBA illustrating this language feature). When someone throws an object such as a ball, as soon as the ball is released the action is completed. Verbs such as talk, walk, write are inherently unbounded, whereby the action does not imply a clear end point. A third category of verbs describes states, such as like, be, and know, which are inherently unbounded. Thus, the AH predicts that learners will first inflect Spanish verbs according to their inherent lexical aspect: Bounded verbs with preterit and unbounded verbs with imperfect and that only later will they learn to inflect unbounded verbs for perfective and bounded verbs for imperfect aspect, in what is called viewpoint aspect. Kissling reports that following C-BLI absolute beginners, produced appropriate viewpoint aspect without first marking lexical aspect, in contradiction to the AH, when completing oral narrative posttest and delayed posttest tasks. Indeed, their performance, especially on the delayed posttest, was closer to native speakers and advanced L2 speakers than is documented for beginning and intermediate learners in the Spanish Learner Language Oral Corpora project. To quote Reference KisslingKissling (2023), C-BLI “helps novice Spanish L2 learners avoid relying on lexical aspect (aspect inherent in verbs and predicates) to motivate their uses of Spanish preterit (PRET) and imperfect (IMP), as the AH predicts they will do in the early stages of learning.”

2.2.4 Phase 4: Languaging

Verbalization reflects the importance of speaking in the mediation of mental activity (see Section 1). We adopt the term, languaging, as suggested by Reference Swain and ByrnesSwain (2006), to capture the process of psychological speech, to distinguish this use of language from verbalization intended to socially communicate with an interlocutor. The significance of languaging is that students begin to relate to their speaking as a teacher relates to it and it produces a self-awareness of what they (think) they know and of what they are doing (Reference Gal’perinGal’perin, 1968, p. 260). Languaging, as a psychological process, begins to move students away from reliance on the external SCOBA and toward reliance on themselves when carrying out an action. Speaking, of any kind, has both a material aspect (sounds, or in the case of writing, marks on a page) and a symbolic aspect. The material side makes it observable and the symbolic side allows us to abstract away from a particular context or object, such as a SCOBA, and in a sense take it with us. It is interesting that the student in the Chinese topicalization study used part of her body, which of course is material, to symbolically replace the rods in the SCOBA – she had virtually transported the SCOBA with her.

There are two subphases of phase 4 that contribute to the abstraction process necessary for internalization. Making one’s understanding and use of a concept comprehensible to someone else relies on the important connection between social and psychological function of speech. Explaining to another enables the other to understand but at the same time it helps the speaker understand something more deeply. Moreover, it is important for teachers to access students’ knowledge of a concept and how they deploy it in communicative activity in phase 5. This subphase is called communicative languaging. The dialogic phase transitions to self-talk, which over time becomes increasingly abbreviated as in the puzzle example presented in Section 1, and is important because it entails the ability to control our mental activity through speech even when someone else is not there to listen. At first the speech is external but eventually, in phase 6, it transforms into inner speech and will no longer be overtly observable.

The first example of languaging is from a student struggling to reconcile what he had previously been taught about Spanish verbal aspect and the new knowledge presented in Negueruela’s class. At the midpoint in the semester, through communicative languaging, the student revealed his problem: “It’s more difficult to speak and rationalize using a certain tense for me, mainly because the reasoning is different from what I’ve been taught in the past. I’m still stuck trying to rationalize it using old methods and it gets confusing sometimes” (Reference NegueruelaNegueruela, 2003, p. 356).

The student’s confusion created by the clash of old and new knowledge is an important component of the developmental process. Later in the semester, the student seems to have resolved the conflict when commenting on the new conceptual knowledge:

When students experience conflicts in their thinking, teachers can play an important role in helping to mediate and resolve the conflict. For Reference VygotskyVygotsky (1987), conflict, which he described as a drama is an important source of development. Without the clash of current and future (new) knowledge there is no motivation for change. However, unless the conflict is made overt, it is unlikely that teachers will even be aware of what is going on. Languaging plays a key function in exposing and resolving the drama. In Section 3, we address ways of mediating students when they experience difficulties using new concepts.

How students react to the languaging phase is also revealing particularly since psychologically oriented speech is not something students have likely experienced through other pedagogical models. One student from Negueruela’s class remarked that “Although sometimes recording myself speak was a bit awkward, I think it was overall extremely helpful. It made me more comfortable speaking and improvising, and it forced me to truly think about the grammar.” (Reference NegueruelaNegueruela, 2003, p. 308). Another student said, “I feel as though with verbalization exercises I not only improved my speaking, but also learned a lot of information about the indicative and subjunctive” (Reference NegueruelaNegueruela, 2003, p. 308).

In a C-BLI study by Reference Yáñez-PrietoYáñez-Prieto (2014) with a more advanced Spanish class, students explained through languaging why they made specific aspect choices when creating personal narratives. Following a presentation on Spanish verbal aspect, including a sequence of SCOBAs (available in the article), the students created a written story on a topic of their choice. They were then asked to explain how they decided to manipulate aspect to create certain temporal perspectives in the story. One student offered the following set of comments:

The student’s comments show a deep understanding of how language can be used creatively to profile events in different ways just as an artist might visually manipulate reality to impart particular impressions on an observer. Clearly, this requires understanding of the concept of aspect well beyond how it is traditionally explained in textbooks through rules of thumb. It is worth noting that the student’s comment is not restricted to Spanish but references “language,” an indication perhaps that she has developed a deeper appreciation of how to create in her native language.

It may seem to some teachers, as Reference Gal’perinGal’perin (1968, p. 259) points out, that engaging students in languaging activity is “an encumbrance in class work.” However, the pedagogical research that he and his colleagues (see Reference TalyzinaTalyzina, 1981) carried out in nearly 800 schools showed that omitting this phase of the model markedly slowed down the developmental process resulting in greater difficulty for students to achieve control over a concept (Reference Gal’perinGal’perin, 1968, p. 259). Languaging is essential for shifting from the social to the psychological function of language. It empowers the individual to engage in abstract reasoning as it mitigates the need to rely on objects, such as SCOBAs, to mediate thinking. The ability to reason about one’s behavior, as illustrated in the above languaging excerpts, is essential for full development.

2.2.5 Phase 5: Performance

The performance phase is open to the preferences of teachers, students, and programmatic policy. The essential point, however, is that without linking conceptual knowledge to concrete goal-directed activity, education falls victim to intellectualism. Reference VygotskyVygotsky (1987, p. 169) stressed that everyday knowledge lacks the capacity for abstraction, meaning that what we appropriate in everyday life is not always open to consciousness and therefore not subject to voluntary control and generalization. On the other hand, he also cautioned, if scientific knowledge developed in school is not sufficiently saturated with concrete activity, the result is “verbalism” [italics in original] (p. 169), or “the mindless learning of words” (p. 170) in which the concept itself is missing. Reference HolzmanHolzman (2009, p. 48) captures the importance of saturating knowledge with concrete activity in the distinction she makes between schooling as “acquisitional” or as “developmental” learning. In the former, schooling produces “knowers,” while in the latter it produces “learners,” by which she means students come to understand that learning is what people do. It is not what people know.

While we recognize that the pedagogical activities such as task-based learning and content-based instruction can be ways of saturating knowledge with concrete action, we believe that drama-based approaches to teaching are especially appropriate for a pedagogical model grounded in Vygotsky’s psychological theory. As we mentioned, drama played a central role in Vygotsky’s thinking about development, most especially when he referenced the work of the great director, actor, and theatrical teacher, Stanislavski as contributing to his thinking about human development in and out of school. In theatrical drama there is a “collision of characters on the stage” (Reference VeresovVeresov, 2010, p. 88), something Vygotsky transferred to what he saw as happening between individuals in the social drama of real life. Veresov provides the example of a debate between two people to illustrate what Vygotsky may have had in mind in conceptualizing development as drama. In a debate the participants generally adopt and argue for opposing viewpoints on an issue. This, Veresov speculates, could result in the participants reflecting on what transpired after the debate is concluded. Perhaps one participant may think that she made a mistake on a particular point during the debate or perhaps she expressed herself in an overly aggressive way. The reflection entails several mental processes: memory of what happened, emotion at having said something wrong and in an inappropriate way, volition, and determining that future behavior must change (p. 88). Consequently, the person changes and in the future behaves in a different way. This is development. In his analysis of development from infancy to adolescence, Reference VygotskyVygotsky (1998) stressed the dramatic collision, or crisis, between children and the specific social reality that surrounds them at each age and how the resolution of the collision results in a transformation, or development of the children. A critical aspect of the collisions is emotional reaction of the individual. Indeed, Reference Vygotsky, van der Veer and ValsinerVygotsky (1994) criticized contemporary psychology for ignoring the intimate interconnection between emotion and thinking. For him, emotion provided the motivation for thinking (and speaking).

During the performance phase, especially early on, students might need to rely on phase 3 SCOBAs to help support their use of a language concept. However, as they gain experience manipulating a particular concept or set of concepts, they are expected to decrease their reliance on the overt manifestation of SCOBAs. As this happens, it is an indication that they are transitioning to internalization – phase 6.

2.2.6 Phase 6: Internalization

In this phase, learners have internalized a concept and no longer need to rely on mediation from a SCOBA, a teacher, or anyone else to support their performance. The individual can access the concept with sufficient speed, although not necessarily as fast as a native user, to participate in spontaneous communicative interactions and/or to produce meaningful written texts. We comment on speed of performance in Section 4. This phase emerges as a result of the activities carried out in phase 5. As learners engage in additional, and more complex, activities they come to rely on themselves as autonomous self-mediated performers and are able to create meanings that reflect their own expressive and communicative goals. This is the phase of inner speech, where planning what to say (or write) is carried out covertly in the individual’s mind before it is executed overtly.

2.3.1 English Particle Verbs

A feature of English that is often difficult for learners to master is the category “particle verbs,” comprised of a verb core followed by a particle such as in, on, over, up, down, and so forth. Examples of such verbs include “run over,” “look up,” “let down,” “make up,” and “single out.” The difficulty learners encounter with particle verbs is their metaphorical use, such as in the sentence “The football team let down their fans by losing the long-awaited match with their cross-town rivals.” Traditionally particle verbs are presented to students as lists that they memorize without much understanding of how to interpret their metaphorical meaning. However, using the concept-based approach of cognitive linguistics, a C-BLI study by Reference LeeLee (2012) and reported on in Reference Lantolf and PoehnerLantolf and Poehner (2014) formulated an effective way to explain and visually represent the meaning of particle verbs. The crucial aspect of the explanation is for students to understand that the metaphorical meaning of these verbs is derived from their literal meaning. The SCOBA presented in Figure 2 captures the relationship between the literal and figurative meanings of verbs formed with the particle “out.”

Figure 2 SCOBA for verbal particle out

(Reference Lantolf and PoehnerLantolf & Poehner, 2014)

The eight sentences below exemplify the various meanings of particles verbs with “out.” The numbers in parentheses indicate which of the four options applies to the sentence.

1. 1. I sleep with my hands out of the blanket. (1)

2. 2. The man went out of his mind with anger. (3)

3. 3. The large dog ran out of the pack and attacked the stranger. (2)

4. 4. The professor singled out the new theory for special attention. (2)

5. 5. The dog jumped out of the window. (3)

6. 6. How did the police figure out who the criminal was? (2)

7. 7. The caterers spread out the food across the table. (4)

8. 8. I filled out my tax forms incorrectly. (4)

2.3.2 Spanish Verbal Aspect

One of the most difficult problems for English speakers to master in a romance language such as Spanish, is verbal aspect when it is expressed in the past tense. Aspect is a temporal perspective to profile events and states. It is a common, if not universal feature, of language. Many languages indicate a wide array of verbal aspects, including English. The problem that English-speaking learners of Spanish face is that in the past tense, a Spanish speaker has the option of marking a past event using either of two verbal aspects – preterit or imperfect. English speakers certainly can construe past events from a different aspectual perspective, but the language does not have clear options, such as specific verb endings, for profiling the distinction.

Briefly, the underlying feature that captures the aspectual distinction is “boundedness.” This is a feature not only of language but also of objects in the world. For instance, water is by its nature unbounded because it does not have a specific recognizable shape. Its shape is determined by the container in which it is put, such as a lake, a glass, a pool, an ocean, a puddle, and so forth. Other objects are naturally bounded in that they have clear limits that we can observe, such as a ball, a fork, an apple, a piece of cake. Events are also naturally bounded or unbounded. The action of “walking” is temporally unbounded because in theory and with sufficient energy one could walk without end. Other events such as jumping, or throwing a ball, are temporally bounded in that they have temporal limits – a natural beginning and a natural end. When someone jumps the event begins and there is a natural end as the person returns to the original position or achieves some other end point, as jumping down the stairs. Grammar, however, allows a speaker or writer to profile what is naturally a bounded event as unbounded and vice versa, an unbounded event as bounded (see viewpoint aspect discussed in Section 2.2.3). The SCOBA in Figure 3 represents the conceptual notion of boundedness. To profile an event as unbounded Spanish speakers use one set of verb endings and to profile an event as bounded they use a different set of endings. Reference Kissling and MuthusamyKissling and Muthusamy (2022) created animated SCOBAs for Spanish aspect available on YouTube.

Figure 3

(a) Boundedness with objects

(b) Boundedness with events

(c) Grammatical Boundedness

3.3.1 Theoretical Basis for DA: The Zone of Proximal Development

The preceding example showed how important information about learner abilities can be revealed when assessments depart from the typical procedure of observing individuals as they independently complete tasks. This innovation in assessment began with two important observations Reference VygotskyVygotsky (2011) reported from his work with children. The first was that some children identified as having special needs and placed in remedial programs responded well to instruction frequently closing the gap with their age peers. This meant that an individual’s performance in school could be changed if appropriate forms of instruction were available. The second observation occurred in his examination of children’s performance on standard tests such as IQ administered upon entering school. He noted that children with initially high scores performed well in school but when retested later some had not improved their scores and that among students who began with lower scores, some performed better at the end of the school year. Vygotsky’s interpretation of this was that the school curriculum provided the challenge necessary for some learners to develop but not others.

Consequently, Reference VygotskyVygotsky (2011) argued that all students would be better served if they were grouped so that instruction aligned with abilities that were maturing rather than those that had already matured. He reasoned, however, that a different kind of assessment was needed: “it is common to think that an independent solution of the problem, without any help, points to the level of the development of intelligence … [but] what is indicative of the child’s intellectual development is not only what he [sic] can do himself, but probably more so what he can do with the help of others” (Reference VygotskyVygotsky, 2011, p. 203) .

Vygotsky then reported piloting-mediated assessment by contrasting two ten-year-old children, both of whom were only able to independently complete tasks designed for eight-year olds. When the children were offered mediation such as reminders, leading questions, a demonstration of principles involved, hints, and feedback, important differences appeared: one child managed to complete tasks at the level of a nine-year-old and the other a twelve-year old. The point was not that the children did better when offered mediation, but that the extent of their improvement was not the same (Reference VygotskyVygotsky, 2011). Their independent performance did not predict how they would respond to instructional support, and it is this responsiveness that is most relevant to future development.

Vygotsky formalized this research in his discussion of the ZPD. The ZPD refers to those abilities that have not yet matured but are still maturing. In most assessments, learners are required to operate without support, and so they rely exclusively on abilities that have already developed. What is revealed by such assessments is the learner’s past; that is, development that has occurred up to the time of the assessment. This, Reference VygotskyVygotsky (1998) insisted, is only part of the picture. In fact, he regarded the ZPD as more important for education because effective teaching should target abilities that are emerging rather than those that have already completed their developmental trajectory. It was for this reason that identifying a learner’s ZPD is of “great practical significance” for education (Reference VygotskyVygotsky 1998, p. 204). Therefore, assessment must be broadened to encompass both the past and future of development as accomplished through learner independent performance on assessment tasks but also how they respond to mediation.

3.3.2 Sandwich and Cake Formats of DA

Vygotsky’s colleague, Reference LuriaA. R. Luria (1961), first shared research on the innovative approach to assessment with an international community of psychologists and educators. The term “dynamic” to describe assessment that captures the ZPD seemingly originated with Luria’s presentation, and soon afterward researchers began to design their own approaches to DA. Given that each approach developed in a specific cultural context, for use with particular learner populations and focused on a distinct set of abilities, differences exist among the approaches. Each is characterized, however, by the inclusion of mediation during the DA procedure as well as analysis of learner responsiveness as part of the diagnosis of their abilities.

Reference Sternberg and GrigorenkoSternberg and Grigorenko (2002) coined the terms “sandwich” and “cake” to designate when mediation is offered, with the most popular format following a three-step process: (1) learners complete the assessment independently; (2) the assessor and learner review the assessment together, with the assessor offering mediation to probe learner understanding of problem areas; and (3) the learner independently completes a parallel version of the original assessment. The three steps together constitute the DA, with the mediation “sandwiched” between two traditional assessments. This format may seem familiar as it resembles the pretest – intervention – posttest design frequently employed in experimental research. As with such research, the sandwich format enables comparisons of learner performance prior to, and following, mediation.

Reference Budoff and HellmuthBudoff (1968) and Reference Budoff and FriedmanBudoff & Friedman (1964) carried out early sandwich-based DAs. At the time, US racial minority children were referred for special education services in schools at much higher rates than their majority peers. The researchers used a sandwich format in tests of general cognitive ability, reasoning that learner responsiveness to mediation could indicate the extent to which they would need special education support before rejoining their peers. They reported that it was possible to distinguish learners whose initial score was very high, those who evidence substantial improvement on the retest, learners who made modest gains, and those who failed to improve. Some DA researchers, beginning with Budoff, have employed the term “learning potential” to express the degree of change in learner performance through mediation, suggesting that high learning potential indicates relatively little instruction will be needed before a learner is able to independently reach the level of performance currently accessible with mediation only. Such information offers one way of capturing the ZPD of individual learners and may be used to inform instruction and other decisions about learners (e.g., program acceptance or placement at an appropriate level of study).

The “cake” format involves a single assessment administration. Mediation is provided following each assessment item or task, thus creating layers similar to the layers of a cake. An advantage of this approach relative to the sandwich format is time efficiency. In addition, Reuven Feuerstein, a leading DA advocate, argued that for learners with a history of academic struggle, excluding an independent pretest avoids reinforcing negative experiences (Reference Feuerstein, Rand and HoffmanFeuerstein, Rand, & Hoffman, 1979). Feuerstein originally formulated his DA model with children who had survived the Holocaust and were relocated to Israel. He refined it through decades of clinical work with children facing extreme developmental difficulties, many of whom could not have their needs met in local schools. Excluding a measure of independent performance means that the cake format does not permit pre- and post-mediation comparisons. Instead, attention is given to the amount and quality of mediation as well as to learner success in identifying and overcoming their mistakes.

3.3.3 Interventionist and Interactionist Approaches to DA

Just as the timing of mediation varies in DA, so too does the quality of mediation available to learners. DA generally begins with implicit and shifts to more explicit mediation depending upon learner responsiveness. The logic here is that if a learner is offered an explicit form of mediation, such as stating the nature of the problem and the principle to be followed to overcome it, we have sacrificed the nuance of the diagnosis, that is, the ability to differentiate among learners according to their ZPD. In fact, if obtaining insights into a learner’s ZPD is not the goal, then we could simply correct their errors, as some SLA researchers have suggested, and not engage in mediation at all. However, this would mean, for instance, that the L2 learners such as Nicole and Chris, might be regarded as having reached the same developmental level because neither learner succeeded in completing the assessment task independently.

Although all DA approaches organize mediation from implicit to explicit, some standardize mediation in advance of procedures, while others favor an open-ended approach in which mediation emerges through dialogue with learners. Reference Lantolf and PoehnerLantolf and Poehner (2004) label these interventionist and interactionist approaches, respectively. In interventionist DA, each mediating “move” is arranged as a protocol or script that is followed in precisely the same manner with every individual. Standardizing the procedure facilitates interpretation of assessment outcomes, whereby it can be reported that one learner required two mediation moves, while another needed four to complete the same task. This can also be formalized in how scores are reported in interventionist DA. For example, scores can be adjusted to award points according to the quantity of mediation a learner requires such that a learner who completed an assessment task worth five points receives three points if two mediation moves were needed, while a learner needing four moves would receive one point.

Advocates of interactionist DA worry that following a script might miss important diagnostic information. In interactionist DA, the assessor is free to ask questions that probe learner understanding even when a correct response is submitted. The assessor can also offer various forms of support when errors occur and may introduce hypotheticals beyond the current task to verify learner reasoning. Such mediation is difficult to quantify; therefore, outcomes of these procedures are frequently reported in the form of qualitative profiles of learner performance and the mediation employed. While profiles do not allow for comparisons the way scores do, they do offer in-depth information to guide instructional decisions. For this reason, interactionist DA lends itself particularly well to classrooms, tutoring, and clinical contexts where the focus is on obtaining as much information as possible into learner abilities. Interventionist DA, on the other hand, is especially well-suited to more formal assessment situations where priority is given to generating scores.

3.3.4 Reciprocity, Transcendence, and Transfer

If mediation draws our attention to what the assessor does, the other side of that coin is what the learner does. While some researchers use “responsiveness” to describe learner reactions to mediation, others prefer the term learner reciprocity, coined by Reference LidzLidz (1991, p. 10), which encompasses more than correcting mistakes. Lidz discovered that children engaged in behaviors that could be usefully included in diagnoses of development but that went beyond whether they improved their assessment performance as a result of mediation. Reference Van der Aalsvoort, Lidz, Van der Aalsvoort, Resing and RuijssenaarsVan der Aalsvoort and Lidz (2002, p. 122) subsequently proposed eight categories of learner reciprocating behaviors: responsiveness during interaction with the mediator; self-regulation of attention and impulses; affective quality of interaction with the mediator; communication related to shared activity; comprehension of activity demands; use of mediator as a resource; reaction to challenge; and modifiability in response to interaction. The researchers note that there may not be relevant observations in each of these categories, but they encourage practitioners to consider these and perhaps other forms of learner engagement in their diagnoses. For example, a learner who has difficulty regulating impulses may resort to guessing, which complicates interpretations of their performance because it can be difficult to distinguish errors that result from guesses and those that reflect reasoning problems. Similarly, through attention to learner communication during the activity, it is possible to ascertain how they perceive the task and the process through which they arrive at their response. These aspects of learner engagement might become targets for instructional intervention.

Transcendence and transfer are related but not identical concepts emphasizing that the focus of DA must always be learner development and not merely successful task completion. DA should not be considered a form of “training” to help learners become more efficient at completing a particular task. Instead, DA seeks to understand the abilities that learners draw upon as they complete tasks, where they experience difficulties, and the quality of effort needed to promote their development. The abilities in question, while invoked by the assessment task, are not limited to it but should apply to other situations learners are likely to encounter in the future. The term transcendence was introduced by Feuerstein (Reference Feuerstein, Feuerstein and FalikFeuerstein, Feuerstein, & Falik, 2015) to express this idea. For Feuerstein, memorizing the answer to a multiplication problem allows learners to correctly answer that problem, but learning how to multiply numbers prepares learners for any such problem they may face because these abilities transcend the immediate task.

Transfer formalizes this reasoning into DA by including tasks that depart from those learners have previously encountered, introducing new parameters or degrees of complexity. Reference Campione, Brown, Ferrera, Bryant, Rogoff and WertschCampione et al., (1984) and Reference Brown, Ferrara and WertschBrown and Ferrara (1985) pioneered the use of transfer tasks in their DA research involving children’s literacy and numeracy development. The aim of Reference Brown, Ferrara and WertschBrown and Ferrara’s (1985) approach is to generate two axes of learner abilities, one representing how quickly learners grasp new principles and concepts and the other capturing how well they apply those when facing novel situations. To this end, learners first complete near transfer problems that are similar to tasks they have previously worked through with mediation but that require use of the concepts they have learned in new combinations. For far transfer problems, learners need to use similar but slightly different concepts in addition to those they have already learned in order to reach the solution. Finally, very far transfer problems are even more complex and may push learners to use concepts in novel domains. While the details of how transfer is used may differ from one study to another, the principle that we should be able to recontextualize our abilities is in line with Vygotsky’s view of development. Indeed, for instruction to promote development, learners need to continue to be challenged by tasks that push them to extend their development to new and more complex domains.

3.4.1 Processes of Mediation and Learner Reciprocity in L2 DA

Reference PoehnerPoehner (2005) implemented DA with university intermediate-level learners of L2 French that focused on their conceptual understanding and use of the tense–aspect system. Students completed oral narration tasks in which they watched brief videos and then recounted what occurred using French. The mediator used an interactionist cake format that probed the learner’s knowledge of French tense and aspect, specifically passé composé and imparfait, including their reasons for selecting each form when portraying particular events. This initial DA was used to generate profiles of each learner that in turn informed an instructional program focused on the tense–aspect system. The program was organized according to C-BLI principles but was tailored to individual learners based on their reciprocity to mediation. For example, some learners required practice on the formation of the passé composé and imparfait, including irregular forms, others were introduced immediately to the concept of verbal aspect, and still others required instruction on the inadequacy of previously learned rules for aspect use. Following approximately six weeks of instruction, a parallel version of the initial DA was administered along with a near transfer task depicting more complex events and a far transfer task for which the prompt was a written excerpt from a French literary text.

Analysis of video recordings of mediator–learner interactions documented the implicit-to-explicit organization of mediation employed along with a variety of learner reciprocating behaviors (Reference PoehnerPoehner, 2005). These are reproduced in Table 2.

Poehner was able to trace across interactions how much mediational support was needed to resolve particular kinds of problems. Note that the forms of learner reciprocity (right-hand column in the table) do not neatly parallel the mediational hierarchy. Arguably, some of these moves represent greater effort on the part of learners to assume responsibility for completing assessment tasks, but this is not straightforward. For instance, one learner might make a guess and then turn to the mediator for approval, while another might reject an offer of mediation resulting in failure to overcome the problem. Rather than higher or lower reciprocity, these might best be understood, as categories of learner engagement in DA that can provide additional insight into their ripening abilities (see Reference Van der Aalsvoort, Lidz, Van der Aalsvoort, Resing and RuijssenaarsVan der Aalsvoort & Lidz, 2002).

Another early L2 DA study greatly extended how learner reciprocity can be understood. Reference AbleevaAbleeva (2010) administered a DA with learners of L2 French focused on listening comprehension, an especially difficult feature of the language owing in part to the complexities of French phonology. In one-to-one sessions, the mediator listened alongside students to authentic French video texts and offered prompts and leading questions to ascertain where comprehension problems occurred, including not only phonology but also knowledge of vocabulary, grammar, and culture. Reference AbleevaAbleeva’s (2010) participants produced categories of reciprocity similar to those in Reference PoehnerPoehner’s (2005) study, and she reported an additional category, imitating the mediator. This is noteworthy because it recalls Reference VygotskyVygotsky’s (2012) argument that imitation is a driver of development. However, he understood imitation not as simple copying or mimicry but as a phenomenon in which a person understands the intentions of the model, that is, the meaning behind each action. For instance, a person with no knowledge of music will perceive a conductor as moving hands about and looking randomly to different sections of an orchestra. Such a person would be able only to emulate but not imitate the conductor because imitation, in Vygotsky’s sense, requires understanding how every stroke of the baton communicates important meanings to the musicians. Imitating the mediator in Reference AbleevaAbleeva’s (2010) project referred to learner efforts to regulate their comprehension of the video texts, often by repeating the mediator’s utterances.

According to Reference AbleevaAbleeva (2010), learner efforts to revise their comprehension of the texts in response to specific forms of mediation could be classified as progressive or regressive, thus revealing whether or not the revision displayed improvement in comprehension. Progressive moves include making a correct choice, deciphering a pattern or word correctly, and overcoming a problem, while regressive moves concern failure to make corrections, as when a learner switches from one incorrect response to another. Distinguishing progressive and regressive moves contextualized not only the number of mediating moves individual learners required to work through difficulties but also the effect of each form of mediation on learner performance. This information also revealed that some forms of mediation proved to be particularly useful for certain individuals, which was considered in preparing subsequent instruction.

3.4.2 L2 Group DA (G-DA)

In Vygotsky’s discussions of the ZPD, he characterized it as involving dyads comprised of a child and an adult or a student and teacher. However, recall that he described the ZPD as important for identifying abilities that are still emerging and may be most easily guided by teaching. By grouping learners according to their ZPD, instruction can more effectively target those abilities and promote the development of each member of the group. With regard to assessment, is it similarly possible to introduce tasks that are beyond the present abilities of all learners in a group and to provide mediation to monitor the group’s responsiveness? In other words, can the same principles that guide one-to-one DA be followed for procedures with a group of learners? If so, DA would be much more feasible in contexts where teachers are responsible for dozens of students and may not have the time and resources to engage with individual students for extended periods of time. Exploring this possibility is the goal of group DA (G-DA).

To our knowledge, the first extension of DA to an intact L2 classroom was reported by Reference PoehnerPoehner (2009). Tracy (pseudonym) was an experienced L2 Spanish instructor who designed her own primary school curriculum implemented in fifteen-minute daily lessons. She had become interested in DA after participating in professional development workshops and regarded it as a way to systematize her classroom-based assessment practices in order to track learner progress over time. This was especially important given the large number of students she worked with daily over a short time span. Tracy was concerned that in using interactionist DA, which she found attractive, it would be difficult to keep track of mediation as she engaged with full classes of young children. She therefore opted to use an interventionist approach, scripting in advance of each lesson a set of eight mediating prompts arranged from implicit to explicit. An example of these prompts is reproduced in Table 3.

Tracy sought to optimize students’ engagement with Spanish during their fifteen-minute mini-lessons, and thus frequently made use of games involving the entire class (Reference PoehnerPoehner, 2009). This approach allowed all students to potentially benefit from the mediation provided to individual students. For instance, using the prompts from Table 3, she responded to learners who had difficulties marking Spanish noun-adjective concord. While Tracy’s mediation supported an individual learner, other learners were able to observe the mediational process. Thus, when it was their turn to respond in a game, they were not necessarily beginning at the same point as previous learners had been because they had vicariously worked through the process of selecting adjective forms and observing Tracy’s guidance. Reference PoehnerPoehner (2009) illustrates this with a sequence of learners who each took a turn playing a game that required them to describe an indigenous Peruvian animal. The first player required the most extensive support from Tracy, only completing the task when offered a choice of forms (the sixth mediating move from Table 3). Subsequent players required only the third mediating move and ultimately one learner required no support at all. Viewed collectively, Tracy was simultaneously mediating each learner’s understanding of Spanish noun-adjective agreement even though she interacted with individual students.

A different approach to G-DA was explored by Reference Poehner, Infante and TakamiyaPoehner, Infante, and Takamiya (2018) in an advanced undergraduate L2 Japanese writing course. The instructor, Sayuri (pseudonym), was a native speaker of Japanese and an experienced teacher, who became interested in DA after participating in a professional development workshop. Sayuri reorganized the process approach to writing and revision that she had previously used after learning about the ZPD and DA. Previously, she had relied on several procedures, including one-on-one sessions outside of class to review student drafts and offer feedback, in-class peer review of student texts, and a whole-class lesson on grammar topics appropriate for advanced learners but that did not specifically address struggles the class was experiencing.

Each activity was re-conceived as an opportunity for mediation attuned to learner emerging abilities. The one-on-one session was replaced by an individualized DA session in which she still reviewed student drafts, but instead of explicit corrections, she used graduated prompts, noting the language features that proved difficult for individual learners as well as how much support they required to identify errors. She used this information to place students in groups according to common struggles. Class time previously reserved for peer review was dedicated to small group work. Each group received a tailored packet of short texts containing the error types observed during Sayuri’s mediation in the individualized sessions. The groups worked together to support one another and to collectively revise the errors they identified. Sayuri circulated among the groups to offer additional mediation as needed. Finally, rather than following a predetermined syllabus of advanced grammar topics, whole-class instruction was devoted to those errors that appeared most common in students’ writing as well as errors that involved complex topics that Sayuri believed required explicit and detailed presentation.

According to Reference Poehner, Infante and TakamiyaPoehner, Infante, and Takamiya (2018), the teacher reported that the new DA-informed approach to writing revision enabled her to track learner progress toward targets that she identified for them as individuals even though much of the mediation was focused on the class. The G-DA was successful in supporting teacher and student interaction. At the same time, it is also the case that this was not a context in which assessment results needed to be reported to an external authority for purposes of high-stakes decision-making. In such a situation, there would no doubt be increased attention to generating scores or ratings for individual learners, and this approach to G-DA would perhaps require some degree of standardization of certain activities or procedures. As discussed earlier regarding interventionist and interactionist DA, the advantage for ease of comparison that is afforded by standardizing procedures generally comes at the cost of limiting mediator flexibility and, consequently, the potential to diagnose individual learner development. Balancing these interests is something that anyone pursuing DA must consider, taking account of the goals of the procedure and features of the context.

3.4.3 L2 Computerized DA (C-DA)

The question of whether DA might be efficiently undertaken with large numbers of learners has also led to efforts to computerize mediation in assessment contexts. Among the earliest discussions of computerized DA, or C-DA, were proposals by Reference JacobsJacobs (2001) and Reference Guthke, Beckmann, Lidz and ElliottGuthke and Beckmann (2000). An important feature of both is that the assessment is paused when learners produce errors and is only resumed after brief instruction addressing the problem. In this way, the approach to mediation, while standardized, nonetheless embeds opportunities for teaching that includes tracking when learners require intervention and how effective it is in improving performance on the remainder of the test.

Reference JacobsJacob’s (2001) C-DA model focused specifically on language aptitude among preschool-and school-aged children. The procedure began with videos of puppets introducing vocabulary and morphological patterns of an invented language based on Swahili. Children responded to a series of questions that assessed their ability to infer word meanings and to appropriately inflect words based on what they had learned from the video. When a learner failed to correctly answer a question, the computer program replayed the part of the video in which the relevant information was presented, and the learner was again presented with the question and permitted to formulate another response. If the second attempt was unsuccessful, the same video excerpt was replayed and a third attempt was permitted. If the learner was still unable to correctly respond, the program skipped to the next question. Outcomes of the assessment included the number of attempts learners made for each item, which the researchers interpreted as revealing the extent of their understanding of particular features of the language.

Reference Guthke, Beckmann, Lidz and ElliottGuthke and Beckmann’s (2000) C-DA approach offered an extension of the Leipzig Lerntest, a form of DA Guthke pioneered and employed with learners of various ages and that targeted a range of abilities. While we are unaware of any publication documenting the empirical results of their C-DA work, the authors’ description of how the system functioned is worth considering. The C-DA was designed around constructs that were broken down into several dimensions with pairs of test items targeting each dimension. If learners answered one or both items in a pair incorrectly, the program paused the test and presented a brief instructional module that included an explanation of principles involved and a set of training tasks to practice using the principles. The test then resumed with another pair of items reflecting that same dimension of the focal construct. Reference Guthke, Beckmann, Lidz and ElliottGuthke and Beckmann (2000) explained that this C-DA approach made it possible to identify both the underlying sources of learner difficulty (i.e., the dimensions of the construct that proved difficult for individuals) as well as whether those abilities were in a learner’s ZPD as revealed by their post-instruction responses.

Reference JacobsBoth Jacobs’ (2001) and Reference Guthke, Beckmann, Lidz and ElliottGuthke and Beckmann’s (2000) C-DA approaches entailed a single form of mediation, namely an instructional presentation, that could be made available to learners as needed. More recent efforts to design L2 C-DA have favored graduated mediational prompts in which a set of suggestions and hints are offered sequentially with the number of prompts learners require indicated in a final report. The first such project was Reference Poehner and LantolfPoehner and Lantolf’s (2013) use of C-DA to diagnose reading and listening comprehension among university learners of Chinese, French, and Russian. Both reading and listening comprehension were further divided into sub-attributes that included knowledge of lexis, morphosyntax, and culture, with phonology added for listening comprehension. Written and aural texts like those commonly used in other standardized language tests were selected, and multiple-choice test items were designed to accompany the texts. Each item targeted at least one sub-attribute. The C-DA system in each of the three languages included a set of mediating prompts arranged from implicit to explicit for each test item. According to Reference Poehner and LantolfPoehner and Lantolf (2013), the multiple-choice format allowed learners repeated attempts to answer each question. When a learner’s initial response was incorrect, the most implicit prompt was given and the learner was permitted a second attempt. This procedure continued until either the learner responded correctly or, if the learner’s fourth attempt was unsuccessful, the correct response was revealed accompanied by an explanation. The reason for including an explanation of the solution was that it constituted an additional opportunity to mediate learner comprehension that could impact their performance on the remainder of the test. In fact, even when a learner selected a correct response, an explanation could still be accessed before moving to the next test item in the event the learner had guessed or was uncertain and wished to verify their reasoning.

A challenge that this C-DA program sought to address was how to represent the insights that the tests yielded regarding learner abilities, including their ZPD. A weighted scoring system was devised in which a mediated score was automatically calculated for each test item that ranged from four to zero points depending on the mediation needed by each examinee. At the end of the C-DA, a learner profile was generated that included: an overall mediated score; an overall actual score (assigning either the full four points for an item or no points); a breakdown of scores for items targeting the same sub-attribute; and a learning potential score or LPS. The LPS follows a proposal by Reference Kozulin and GarbKozulin and Garb (2002) to calculate a single overall score that takes into account both a learner’s independent performance (actual score) and ZPD (mediated score). The formula for determining an LPS that Reference Poehner and LantolfPoehner and Lantolf (2013) adapted from Kozulin and Garb is LPS = (2 × mediated score − actual score)/maximum score.

While revisions to this formula have been recently proposed (see Reference Sun, Peng and WangSun, Xu, & Wang, 2023), the overall idea of situating the difference between the mediated and actual score in relation to the maximum possible score on the test has remained consistent. The convenience of producing a single score that initially motivated Reference Kozulin and GarbKozulin and Garb (2002) to pursue an LPS, however, requires some caution. As Reference Poehner and LantolfPoehner and Lantolf (2013) explained, a more nuanced account of learner abilities, precise areas of difficulty they experience, and their ZPD requires information that is obscured if only one score is reported. It is for this reason that their C-DA program included LPS as one component of an overall learner profile.

Reference QinQin (2018) designed an interesting extension of this C-DA model in her work with university learners of L2 Chinese. The listening and reading comprehension constructs that were targeted in Reference Poehner and LantolfPoehner and Lantolf’s (2013) approach meant that successfully responding to test items involved identifying details from the written or aural texts. Qin took this a step further by assessing learner comprehension of implicature, that is, not merely their understanding of an utterance but what pragmatic meaning is likely implied by the speaker. Qin’s C-DA program included recordings of Chinese speakers responding to invitations, suggestions, and requests. In some cases, these were accepted by the speakers while in others they were refused, but how this act was accomplished varied greatly. For instance, a dialogue in which one speaker announces an impending visit to the city where the interlocutor resides and asks whether he could stay in her apartment is met with a response that the apartment is small and cramped. To successfully answer the test question, listeners must comprehend not only what is said but also its implicature. Consequently, the mediating prompts not only needed to draw learner attention to relevant portions of the aural text, they also needed to guide efforts to construct an accurate interpretation. The construct of implicature comprehension is complex, especially given that Chinese cultural norms differ from Western norms, and these differences became the focus of instructional intervention.

A third L2 C-DA model is Reference LeontjevLeontjev’s (2016) ICAnDoIT system. Recognizing the potential for students to guess the correct response to a multiple-choice item, especially when each additional attempt reduces the degrees of freedom, Leontjev prepared sets of parallel test items for each construct to be assessed and permitted learners only a single attempt at each. If a learner selected the right option for the first item in the set, the message Correct! appeared, and the next item was presented. If, however, the learner answered incorrectly, an implicit mediating prompt was given before the second item appeared. If the learner answered this second item correctly, this was conveyed but if an incorrect response was selected a second, more explicit prompt was given. A total of five prompts for each set of test items was included in the ICAnDoIT system, and since no prompt was provided unless a learner made an incorrect choice, some learners completed the entire set of items without any prompts, while others required the full five. While Leontjev’s C-DA model did not automatically calculate an LPS, the fact that each set of items was linked to a specific feature of English grammar (e.g., Wh- question formation) meant that by tracing the number of prompts learners received per set, it was evident which areas of grammar were most challenging for each individual.

4.3.1 Reactions to BFM

In this subsection, we present the reactions of teachers and teacher educators who have experienced the BFM and who have implemented C-BLI in their own classrooms. The samples are drawn from Reference Esteve, Lantolf, Poehner and SwainEsteve (2018) and Reference Lantolf, Esteve, Sato and LoewenLantolf and Esteve (2019).

C-BI has led me to read some new theoretical books and to revisit handbooks on general linguistics and on the language I teach. It has also led me and my students to challenge well-established beliefs about the language we all are learning (for instance, that heavily colloquial language is easy to deal with) … I feel ‘I am in process’, like my own students. I experience a developmental process running parallel to theirs and find myself being surprised and/or learning alongside them … I believe learners gradually develop a more global view of language and become increasingly aware when they need to make appropriate adjustments. They have a deeper understanding of text genres and their features and are stepping into a less ‘fragmented’ dimension of the language they are learning. (Reference Lantolf, Esteve, Sato and LoewenLantolf & Esteve, 2019, pp. 43–44)

The teacher is clearly rethinking, along with her students, about what language is in general and what the specific language in the teaching–learning experience is. She is willing to investigate what research has uncovered about the nature of language, which is an important transformative step away from traditional beliefs about pedagogical rules. She is learning something new about language along with her students, and they are moving away from a piecemeal understanding of language and toward a more coherent understanding based on language as genre.

I feel my professional development as a teacher involves turning into a language researcher and providing my students with the best possible framework for them to become informed learners able to grasp and handle language as a structured whole. For me, such a framework is none other than C-BI [sic], as it clearly helps learners become language researchers as well through a translinguistic approach leading them to develop their plurilingual competence as a way to enhance their communicative competence (Reference Lantolf, Esteve, Sato and LoewenLantolf & Esteve, 2019, p. 46).

Not only does this teacher engage in research as a means of improving her instructional practice, but because of their experience of language learning through C-BLI, her students also become researchers, not so much as engaging in formal research but to the extent that they explore the similarities and differences among the languages they know and are learning. Most students in the Barcelona regional schools are bilingual speakers of Spanish and Catalan and they also study at least one additional language in school.

In preparing classes and selecting texts to be worked on, I can address language complexity without leaving any aspect unconsidered … Through CBI [sic], I can see beyond purely formal mistakes related to grammar or vocabulary and spot the students’ creative use of language. This use accounts for their transformation process, i.e., for the extent to which they have gained control over the concepts dealt with, even if the linguistic elements used to express such concepts are not totally right (Reference Esteve, Lantolf, Poehner and SwainEsteve, 2018, p. 499).

This teacher clearly values creativity on the part of students, something that is central to Reference VygotskyVygotsky’s (2004) educational agenda. In this regard, recall the study by Reference Yáñez-PrietoYáñez-Prieto (2014) on promoting creative use of L2 Spanish. She also seems less concerned about selecting texts for her students because they might contain language features that they, or perhaps even she, had been unable to deal with.

The work with SCOBAs involves procedures that let emerge [sic] the teachers’ own previous knowledge and professional experiences. It seems to naturally connect what is being learnt with each teacher’s own background. This work represents a way, enjoyable for me and exciting for teachers, to ensure that information which might just have been merely transmitted gets actually transformed into knowledge … (Reference Esteve, Lantolf, Poehner and SwainEsteve, 2018, p. 501).

Clearly for this teacher educator the outcome of the BFM has been very positive, not only in terms of motivation, but also in terms of how teachers think about language and how they are able to integrate conceptual knowledge into their instructional practices, as well as with regard to greater coherent and less idiosyncratic thinking across teachers as a group. The dialectical interaction between OBAs that teachers entered the program with (original pre-understanding) and SCOBAs (depiction of theoretical knowledge) has resulted in a deeper understanding of language and its teaching.