Chapter 14- The Interaction Hypothesis: Why You Shouldn’t Learn Languages Alone

Wikis > Chapter 14- The Interaction Hypothesis: Why You Shouldn't Learn Languages Alone

The Interaction Hypothesis: Why You Shouldn’t Learn Languages Alone

By Chelsea Kung Chi Yee, Connie Soon Kang Ning, & Seth Lim Jun Hua; AY2016/17 Semester 1.


Hi there! Welcome to our WikiChapter. Within this chapter, you will find information about the Interaction Hypothesis. In particular, you will be able to learn about the following sections:

  1. What is the Interaction Hypothesis?
  2. Why the Interaction Hypothesis?
  3. The Interaction Hypothesis in Action: Today
  4. Limitations & Criticisms
  5. Going Forward
  6. Conclusion: So why shouldn’t I learn languages alone?

Navigating this chapter is easy: although these sections were written to be read in order, feel free to click on the links above to get where you want to be!

By the end of the chapter, we hope you would’ve learnt more about the Interaction Hypothesis, and more importantly, why you shouldn’t learn languages alone. Good luck!


Chances are, you’ve come across a language before, that you didn’t grow up speaking. Maybe, at some point, you might want to, or possibly even have to, learn one of these languages. What would be the best way to go about learning an entirely new language? And how would such a method work? Is there even a best method— one that would maximise learning for the effort put in?

Scientists propose that the answer to these questions may lie in the subtle interactions that second language (L2) learners engage in— in particular, interactions involving use of one’s L2. An integral part to communication, interaction through conversation certainly seems helpful in learning a second language, but to what extent? Over the years, various studies have attempted to answer this question by examining the role conversations play in language learning. Through participating in L2 interactions, learners have been found to gain access to conditions known to bolster language learning— conditions such as processing comprehensible input (e.g. Krashen, 1981, 1985), interactional restructuring with focus on form (e.g. Long, 1983, 1991), and production of modified output (e.g. Swain, 1985, 1995). (Note to reader: These terms will be defined later in the chapter. Don’t let them scare you yet!) Put simply, interactions serve as a major source of language data, augmenting the learning process by encouraging the active processing and generation of original content in one’s L2. Through interactions, participants get to practice their internal language mechanism— governing syntax, semantics and structure— and in so doing, get more proficient with each interaction.

There continues to be an increasing amount of evidence in favour of the critical role that interactions play in the language-learning process (e.g. Gass & Mackey, 2007; Leslie, 2015; Ognio, 2008, 2012; Yang, 2012)— enough, even, for linguist Allwright (1984; 156) to refer to them as a ‘fundamental fact of pedagogy’ in L2 education. With a strong case in support of this pivotal role, various hypotheses have since attempted to synthesise these diverse findings into a single, coherent explanation. In particular, the Interaction Hypothesis (IH) has proven itself to be amongst the more robust of explications today.

(return to top)

What is the Interaction Hypothesis?

So what exactly is the Interaction Hypothesis?

The Interaction Hypothesis (or IH, for short), is a theoretical account of second-language acquisition (SLA), which attempts to explain the role of interactions in the language learning process. As opposed to internally-driven acquisition approaches, the IH sits in line with a socio-interactionist approach, which emphasises the influence of the environment in which a learner is engaged.

Nowadays it is believed that language is developed through interaction and negotiation of meaning. ” (Ghaemi & Salehi, 2014; p. 25)

Before going into more detail, however, it is important to first be familiar with the four constructs the IH comprises, Input, Interaction, Feedback and Output. Gass & Mackey (2007) describes these constructs as non-distinct, due to the absence of specific boundaries which divide them clearly one from the other. In fact, each of the four constructs are interlinked in some way, as will be explained below.

Input, Output, Interaction, Feedback

  • Input/Output

    Described as Gass & Mackey (2007; p.177) as the “sine qua non of acquisition”, input refers to any of the linguistic forms received by the learner. Put simply, input can be broadly defined as information received by the learner, from an external source. Output, on the other hand, refers to the linguistic forms produced by the learner— essentially, internally-generated replies to the other conversational party (or parties) in an exchange.

    Traditionally, these two constructs were, by definition, restricted to verbal, face-to-face conversations, ideally with native speakers of the L2. In more recent developments of the IH, however, this narrow definition has since been expanded to include any form of linguistic data— spoken, written, or typed— from classroom interactions, to online exchanges via social media, to even AI-generated speech.

    In the context of the IH, the interlinked constructs of input and output are viewed as the entities exchanged within an interactional setting, which allows interactions to happen Through analysing the input and output of an individual, it is also possible to gauge their relative fluency and proficiency in the language (think: the conversational segment of oral examinations).

  • Interaction

    Defined as the function that input and output fulfil— interactions can generally be described as negotiations for meaning. These exchanges have an interactional structure; in the context of the IH, this refers to the manner in which information is exchanged during an interaction between parties. In this structure, modification techniques such as clarification and repetition are able to come into play to facilitate negotiations for meaning. Through a successful negotiation, learners are able to determine crucial information about their utterances— for example, how standard their phrasing of a sentence was, or whether their understanding of vocabulary matches up with actual use, in context. A successful negotiation can be hindered by a variety of factors, such as language barriers, or cultural differences (especially differences in language use, e.g. across dialects).

  • Feedback

    The construct of feedback is quite similar to input, with the crucial difference being that feedback is received in response to output. Two kinds of feedback can be identified within the context of the IH— explicit feedback such as (but not limited to) corrections, or metalinguistic explanation, and implicit feedback, including negotiation strategies such as clarification requests or comprehension/confirmation checks.

    A further distinction between positive and negative feedback was later proposed. Positive feedback involves (possibly partial) agreement with the learner’s input (e.g. discussion where you expand on the ideas expressed in the input), whereas negative feedback involves rejection (eg. a teacher correcting the grammar of a student in the language classroom) of the learner’s input.

Together, these distinctions allow one to dissect the complicated nature of interactions into observable, explainable parts.

(return to top)

Development of the Interaction Hypothesis

The IH today exists as a collective effort of various authors, developed over the years. Notably, rather than proposing entirely new ideas each time, each iteration of the IH has built upon previous versions, adding on key ideas and theories to augment existing framework.

  • Krashen (1980): the Input Hypothesis

    The IH is largely attributed to have found basis in Krashen’s early work— in particular, the Input Hypothesis, published in 1980. In that paper, Krashen put forth the notion of comprehensible inputthe idea that understanding the input that one encounters is crucial to learning. In fact, Krashen considered comprehensibility of the input to be so crucial, that he proposed learning would not take place unless learners are able to process meaning from the input. While this notion of necessity has since been debunked, comprehensible input remains a major part of the IH.

    The idea of interactional restructuring was also brought up in Krashen’s early work. Especially in the initial stages of SLA, a learner’s understanding (and in turn acquisition) can easily be compromised from the existing language barrier. When such communication barriers happen, modifications are made to the interactional structure, with techniques such as repetition, clarification checks and comprehension checks. Over the course of the interaction, these techniques are in a bid to facilitate the conversation. Such techniques are usually absent when mutual understanding is not inhibited by issues such as language barriers. This is also an example of feedback.

    Modifications to the interactional structure of conversations which take place in the process of negotiating a communication problem help to make input comprehensible to an L2 learner.

  • Long (1983): Interactional Restructuring

    Also emphasising understandable interactions in context, Long’s version of the IH builds upon Krashen’s idea of comprehensible input. In interactional contexts (two-sided conversations involving at least two parties), there is an exchange of information happening. Interactions can occur in varying contexts and formats. Comprehensible input thus builds up an individual’s ability to understand information across varying channels and expressions, which is key in achieving true acquisition of a language (rather than being restricted to just textbook knowledge). The exchange of information promotes interactional restructuring. The amount of comprehensible input determines the rate of acquisition.

    For example, sound discrimination is only applicable in real life where different people express the same phonetic article with varying articulations.

  • Pica (1987): Social status & Feedback

    Pica’s understanding of the IH concentrates on the idea of ease of conversations. Social status differences give rise to disparity-triggered social implications (e.g. people of lower status are discouraged from or are not allowed to question and make suggestions to people of higher relative social status).

    The equality in social status between interlocutors allows for possible use of interactional modifications such as questioning and suggestion as people are able to make use of these tools without restrictions of possible social repercussions. Comparable social status or social roles thus give rise to increased opportunities for interactional restructuring.

    Pica’s emphasis is thus on status-induced opportunities for conversation and feedback.

  • Ellis (1991): a Critical Review

    In 1991, upon critical review of the previous versions, Ellis (1991, p. 36) proposed yet another revision to the IH. As Ellis contended in his version, the claims in his revised version are intentionally crafted to be weaker than those identified in the original IH. He emphasised that the advantage of his revised version is that it is possible to see how the hypothesis can be empirically tested.

    Ellis summarises his proposed IH as such:

  1. Comprehensible input facilitates SLA but is neither necessary nor sufficient.
  2. Modifications to input, especially those which take place in the process of negotiating a communication problem, make acquisition possible, providing that the learners: (a) comprehend the input, and (b) notice new features in it, and compare what is noticed in their output.
  3. Interaction that requires learners to modify their initial output facilitates the process of integration. “

Commenting on his revised revision of the IH, Ellis (2008; pp. 257) wrote:

…with its emphasis on the contribution of negative feedback and modified output as well as comprehensible input and its recognition that interaction works by connecting input, internal learner capacities, and output via selective attention is obviously a major advance on the early version”

At Present: the IH, defined

While many internal definitions have since been broadened, the IH has remained largely unchanged since Ellis’ 1991 revision, with the exception of minor additions to strengthen and reframe the focus of interactions to specific contexts. In particular, the later version of the IH is viewed to be closely related to another type of instruction ‘Focus-on-Form’ where attention to form is highlighted and derived from the performance of meaningful communication tasks. (Ellis, 2008). Additional influences of psychology have brought the notions of selective attention and noticing into the picture, in order to explain some of the IH’s inner workings.

At its core, the IH still proposes that SLA is facilitated by understandable input and informational exchanges, through interpersonal interactions between speakers. Interactions are, of course, still ascribed a major role the SLA process— besides being a significant source of language data, it is also through the active participation in interactions, that much of the learning occurs.

Empirical Basis for the IH

Any robust hypothesis should be able to be proven, or falsified. The IH is no exception. 

Various studies have yielded results which support the claims established by both Ellis as well as other key figures in the development of the IH. In a study carried out by Loschky in 1994, it was concluded that short-term interactional modifications did have an impact on participants’ comprehension of Japanese vocabulary items. Interactions in the form of conversations with interlocutors would thus be useful in the long-term understanding and acquisition of vocabulary items and in turn the language itself.

This is further proven by a study by Ellis, Tanaka and Yamazaki in 1994 where the role of interactional modifications was further amplified with the revelation of better comprehension and greater vocabulary acquisition as a result of intentionally modified input in comparison to premodified input. The relationship between interaction, comprehension and SLA is thus established as a relationship where comprehensible and comprehended input is important and beneficial for SLA.

Additional research by Mackey (1995, 1997) also examined the effects of varying types of input and interaction on the short-term development of question formation. Negotiated interaction with active participation was found to positively impact the development of the ability to form questions, unlike situations where interactions were merely observed, with no participation in the course of interaction. This is evident of the gist of the IH, where participation in interactions (i.e. a two-way communication process) is beneficial to the language acquisition process in general.

Results of the experimental studies mentioned above have sufficiently proved and argued for the general idea of the IH, where comprehensible input and interaction is highly beneficial to an individual’s SLA process. However, as with all hypotheses and theories, there are still limitations in terms of scope of study and possible confounding factors, especially when it is not considered ethically possible for studies to provide a conflicting point of view to IH through deprivation of interaction in the language acquisition process. Such limitations will be further explained and highlighted in a separate section later on in the WikiChapter.

(return to top)

Why the Interaction Hypothesis?

Arguably, language is learned for the ultimate purpose of facilitating conversation. In order to communicate efficiently, interaction is highly important and useful. In the current trends of language studies, it has been identified that learning is most efficient when applied in the context of social usage— in other words, through interpersonal interactions.

It is now generally accepted that verbal interactions are a central platform for second language learners to receive and exchange information for L2 acquisition. This is evident in both naturalistic and classroom settings, where the value of face-to-face interaction is clear for the former, although less prominently so in the latter. After all, learners in classroom settings also engage in activities that involve much less interaction, such as reading and writing. However, it is undeniable that the teacher’s role in the classroom can be gainfully regarded as an interaction that provides learners with opportunities for learning. The study of the relationship between interaction and L2 acquisition, therefore, constitutes one of the main ways in which the IH can inform pedagogy.

Since its early days, trends in second language education has come a long way— from routine drills and constant practice of structural grammar and vocabulary, to the current time, where pedagogies encompass real-life situations to ensure relevance (e.g. context-dependent learning, including English for Specific Purposes, or English for Academic Purposes). Personalisation of information and knowledge has been shown to be important for the efficient absorption of information— it is with this context that we proceed to our next section.

(return to top)

The IH in Action: Today

Now that we understand what the Interaction Hypothesis is all about, how does this apply to SLA in the today’s world? Traditional conversational contexts have already been extensively considered; a quick search on Google can easily offer you a sound understanding of early research on the IH, and more. Instead, here, we consider 4 less-considered modern-day contexts in order to provide insight on how the IH has since been applied in today’s Internet-inundated age.

Let us start by considering interaction in the virtual world. When considering today’s context, one defining feature would probably have to be reliance on technology, and in particular, the internet. This increase in the use of online media in language learning seems to have triggered a shift from a conventional teacher-centred environment, to one that is more learner-centred. In the field of language learning, this could potentially translate into a more positive impact as a whole.


Using electronic mail (e-mail) as a platform for language learning means that learners and teachers can have interaction anytime and anywhere, as long as they have access to a laptop or mobile phone, and a working internet connection, and this also allows the teacher to give feedback to the student, even outside of classroom hours! The email is primarily used to work on the writing skills of learners, and is described as a “dialogue journal” between learners and teachers, and is useful as it is essentially a written tool to convey ideas and convert writing to social activities. Beyond acting as an additional medium between teachers and learners, teachers can also gather individual students, both second language learners of a language and native speakers of the same language, to facilitate communication between them and this would give the learner a greater incentive to learn the language too, in order to be able to communicate well enough with the native speaker (as real world interaction). At the same time, while learning a new language, this also promotes a stronger cultural understanding for the learner.

Online forums

The use of online forums (e.g. learner management systems like NTU Learn), as a space for interaction and discussion among second language learners in the classroom has been increasingly common, for a simple reason which is that there is much more participation in online discussions due to a higher confidence, and perhaps hiding behind a screen and the keyboard takes away the fear of being wrong, or humiliation. This is helpful to those who are shy or uncertain of how to express their ideas vocally. By providing a safe entry point, they are allowed to participate and express themselves in a comparatively more organised manner. Here, the learning process is not just limited to classroom hours where students have to be in a physical space together, which makes it more convenient especially for learners who are full-time students/employees/otherwise preoccupied. However, the learners are not just accountable for their own learning, but also responsible to contribute to the success of the course objectives.

That being said, we also recognize the potential disadvantages and limitations of virtual language learning, which highlights that this might not be a viable medium for all second language learners. First, there is the possibility of learner frustration and confusion, and this might be hard for the teachers to detect if the learner themselves do not voice out. Secondly, there will be a higher student attrition rate, and the need for greater discipline, motivation and commitment from the learner. Also, with this virtual interactions, the time for interaction in traditional classrooms for “real learning” would lessen.

What about social media and gamification in virtual language learning- paving the way forward?

Social media

Carmean and Haefner (2002) explained that social networks are applied online technologies which make the second language learning more social, and enjoyable with minimal stress. Social media such as Facebook, in this case, are used by teachers to encourage interaction and discussion between students in the classroom, as well as with other native and non-native speakers of the language they are learning. With Facebook being a platform that is so easily accessible (1.71 billion active users worldwide), it is also a medium for second language acquisition, where the individual may not be learning the language formally in a classroom setting, but through these interactions with native speakers, they pick up incidental vocabularies and eventually, the language. While the language used may not always be structurally correct and more conversational (informal), it is through more and more interactions, where the individual observes how the language is used by native speakers and through corrections, they acquire more of the language than say, through watching television drama series in the language they are interested in acquiring.

Online games

This transition from the reading and writing grammar-based instruction towards one that is more functional and communicative-based is because educators and and theorists alike have found that second language learners were not able to use the language accurately and efficiently when needed (Tusing & Berge, 2010). Online games are another available resource which can be used to enhance the learning experience of second language learners, where they can practise grammar and vocabulary learnt, and also have interaction with both characters in-built in the game, as well as other real world learners accessing the platform. One such example would be Second Life, a tasked-based game where players would get to explore a virtual world, complete with fictional characters and stages where interaction is needed in order to move on to in the game. The benefits of using such a resource would be that players are intrinsically motivated to play, and the game provides a platform for trial and error play, where users will not be afraid to get the wrong answer, but instead would want to keep trying till they get it right!

With the gradual increase in the demand for personalisation of language-learning processes (Yang, 2012), the IH offers insight on personalisation through designing the curricula for language acquisition around interactions. Once a seemingly distant dream, the incorporation of SLA theories like the IH would certainly result in transformations of  learning space, and pedagogy both offline and online. An optimum learning environment fitted to each learner’s preference, and suited for maximum efficiency of language acquisition, can now be considered merely to be a work in progress.

(return to top)

Limitations and Criticisms

Despite its vastly comprehensive scope, the IH is admittedly unable to cover all elements of the SLA process, instead choosing to focus primarily on the 4 constructs of Input, Interaction, Feedback and Output. Yet, it is this focus that might limit the IH. Interactions, though significant, only make up a portion of the second language learner’s language experience; conceivably, failing to considering other parts of the SLA process could leave room for confounds in the process of interaction that might affect the language learning process.

Furthermore, there are areas of the IH that might prove more theoretical than practical. An example would be the notion of negative feedback. Proposed to be a significant driver of learning, negative feedback allows a learner to gain awareness of their responses to input. Yet, it is unlikely that negative feedback would be found outside of the language classroom. Examples of negative feedback include correcting ungrammatical usages, but explicitly doing so in regular everyday interaction would not seem socially acceptable in the slightest. A similar instance can be found in the negotiation process— in the very beginning stages of learning a new language, a limited vocabulary would likely dictate difficulty even in expressing oneself in complete, grammatical sentences, limiting the meaningfulness of utterances produced. On the other hand, advanced learners might not gain any useful linguistic knowledge when interacting with less advanced learners. This asymmetry in learning is an aspect not completely accounted for in the IH.

Caution must also be taken in applying findings of the IH to the classroom. As Gass and Mackey (2007) caution, the IH describes the learning process, not the teaching process. As such, applying findings from the IH to pedagogical choices would require extra attention. Without being sure of the potential advances in learning, designing entire language courses centered around the interactions that students would potentially experience, might not be the best idea as of yet.

(return to top)

Going Forward

Despite these limitations, the IH has managed to be the focus of a vast number of works (e.g. Ognio, 2012; Tusing & Berge, 2010; Yang, 2012). Given the prevalence of interactions in the language acquisition process, no theory of SLA would be able to be complete without accounting for the role interactions play. To date, the IH remains as one of the most comprehensive hypotheses to that measure (Ghaemi & Salehi, 2014). With the amount of support it has garnered over the years, added to the potential for development, interest in the IH certainly seems set to grow.

Worthy of note would be the IH’s progress towards, possibly, both a model and a theory. In its current form, the IH provides a comprehensive account of the processes involved in interaction. This is characteristic of models. At the same time, in seeking to explain the relationship between interaction and learning, the IH draws links to psychological concepts such noticing and attention. This is characteristic of theories. Given enough time and sustained interest, it would certainly be interesting to follow the development of the Interaction Hypothesis.

(return to top)

Conclusion: So why shouldn’t  I learn languages alone?

The reasons are simple, and numerous too. Listed here are our top 3 reasons why we think you shouldn’t (they’re all mentioned in the chapter too; see if you can remember which sections they’re from)!

  1. Having conversations with others not only gets you more exposure to your L2, but also serves to integrate the necessary information required for acquisition. Talking to yourself is much more boring, and might not achieve the same effect.
  2. Participating in interactions are shown to improve memory for both vocabulary AND grammatical functions. Not participating in interactions means you miss out on these potential gains.
  3. With resources and communities so freely available online, it might actually take more effort to keep your distance from a community of L2 learners just like you.

If none of the above reasons appeal to you, you can at least take heart in the fact that it is much more enjoyable to learn a language with a friend, than all alone. Having a buddy always helps!


(click to expand/collapse)

Bahrani, T., Tam, S. S., Nekoueizadeh, M. (2014). Second Language Acquisition in Informal Setting. Theory and Practice in Language Studies, 4(8), 1714-1723.

Derakhshan, A. & Hasanabbasi, S. (2015) Social Networks for Language Learning. Theory and Practice in Language Studies, 5(5), 1090-1095.

Ellis, R. (1991). The interaction hypothesis: A critical evaluation. Paper presented at the Regional Language Center Seminar, Singapore, April 22-28, 1991.

Ellis, R. (2008). The study of second language acquisition (Second Ed.). Oxford: Oxford University Press.

Ellis, R., & He., X. (1999). The roles of modified input and output in the incidental acquisition of word meanings. Studies in Second Language Acquisition, 21, 285–301.

Gass, S. M., & Mackey, A. (2007). Input, interaction, and output in second language acquisition. In B. VanPatten and J. Williams (Eds.), Theories in second language acquisition: An introduction, pp. 175-199.

Gass, S. M., & Mackey, A., Pica. T. (1998). The Role of Input and Interaction in Second Language Acquisition. The Modern Language Journal, 82(3), 299-305.

Ghaemi, F., Salehi, N. (2014). Interaction Hypothesis: A Comprehensive Theory of SLA? International Journal for Teachers of English, 4(4), 23-33.

Krashen, S. (1981). Second language acquisition and second language learning. Oxford: Pergamon.

Krashen, S. (1982). Principles and practices in second language acquisition. Oxford: Pergamon.

Krashen, S. (1985). The input hypothesis: Issues and complications. London: Longman.

Krashen, S. (1989). We acquire vocabulary and spelling by reading: Additional evidence for the input hypothesis. Modern Language Journal, 73, 440-464.

Leslie, C. (2015). Humour in Peer Interaction in the L2 Classroom. e-TEALS: An e-journal of Teacher Education and Applied Language Studies, 6, 51-67. New University of Lisbon.

Long, M. H. (1991). Focus on form: A design feature in language teaching methodology. In De Bot, K., Ginsberg, R., Kramsch, C., Foreign language research in cross-cultural perspective, pp. 39–52. Amsterdam: John Benjamins.

Long, M. H., & Robinson, P. (1998). Focus on form: Theory, research and practice. In C. Doughty, & J. Williams (Eds.), Focus on form in classroom second language acquisition, pp. 15–41. Cambridge: Cambridge University Press.

Loschky, L. (1994). Comprehensible input and second language acquisition: What is the relationship. Studies in Second Language Acquisition, 16, 303-323.

Ogino, M. (2008). Modified Output in Response to Clarification Requests and Second Language. Hamilton, New Zealand: The University of Waikato.

Ogino, M. (2012). Modified Output, Clarification Requests and Developmental Progress of Learner Language: the Case of Negation of Adjectives in L2 Japanese. New Zealand Studies in Applied Linguistics, 18(1), 5–20.

Pica, T. (1987). Second language acquisition, social interaction and the classroom. Applied linguistics, 8, 1-25.

Pica, T. (1989). Classroom interaction, participial and comprehension: redefining relationships. Papers in Applied Linguistics, University of Alabama, 1,1-36.

Pica, T. (1994). Research on negotiation: What does it reveal about second-language learning conditions, processes and outcomes? Language Learning, 44, 493–527.

Pica, T, Young, R., & Doughty, C. (1987). The impact of interaction on comprehension. TESOL quarterly, 21, 737-58.

Schmidt, R. (1990). The role of consciousness in second language learning. Applied linguistics, 11, 129-58.

Swain, M. (1985). Communicative competence: Some roles of comprehensible input and comprehensible output in its development. In S. M. Gass, & C. G. Madden (Eds.), Input in second language acquisition (pp. 64–81). Cambridge: Cambridge University Press.

Swain, M. (1995). Three functions of output in second language learning. In G. Gook, & B. Seidlhofer (Eds.), Principle and practice in applied linguistics (pp. 125–144). Oxford: Oxford University Press.

Yang, Si-Ho. (2012). Social interaction in second language learning. English Teaching, 67(1), 131-156.


(return to top)