SLA research is dealt with in detail in DELPHI Module 2, so this section serves merely to underline those aspects most relevant to the development of speaking skills. The teaching of oral skills has evolved a long way since the traditional Grammar-Translation approach which laid no emphasis on being able to speak the language. Krashen's Input Theory, which prompted the first teaching approach to be based upon SLA research, relied on giving students lots of comprehensible input in the TL in order to facilitate the processing and understanding of how a language works (see Module 2, section 2.3.2).

Many consider that this approach failed to emphasize sufficiently the productive language skills. Swain's Output Hypothesis argues that the focus on understanding language fails to give students sufficient opportunity to practise their skills (section 2.3.3; see also Swain, 2000). Creating output pushes learners to process the language more deeply; it also necessarily requires social interaction, which forces learners to negotiate meaning and to seek solutions. As they attempt to produce language, students consolidate their knowledge, but become aware of the gaps within it. They form hypotheses, test them out, seek help, and regulate each other's activity. Output subsequently elicits further input, which will present new challenges for the learner to overcome. See Swain (2000) and Chappelle (1998).

A major challenge for language teaching, especially the development of oral skills, is how to build up linguistic progression in a way that is manageable to the classroom learner. Second language acquisition research based on cognitive psychology sees language learning as a gradual process of skill acquisition, in which declarative knowledge about the language is turned into procedural knowledge of language use in 'real time' (see Johnson, 1996). On this view, learners progress from repetition and copying to the 'online' generation of own language in a process of gradual development. Some would equate this with the well-established pedagogical move from the 'pre-communicative', ie structural and quasi-communicative activities, to the 'communicative', or functional and social interaction (Littlewood, 1981).

One of the pedagogical applications of these insights has been the so-called PPP approach, or

• Presentation: drawing learners' attention to a specific form or structure, usually through contextualized use;
• Practice: teacher control gradually eases and learners work on the particular form, initially in controlled conditions;
• Production: learners engage in open practice, free of teacher control, with the focus on meaning.

Many working in SLA (see, for example, Willis and Willis, 1996) are critical of the proposition that it is possible to order 'chunks' of the language into a syllabus of graded difficulty and that a given item can be learnt and subsequently employed in spontaneous language use within the space of a single lesson or even a few lessons. Even after a particular linguistic item appears to have been taught successfully (ie learners have jumped through all the hoops as intended and have used a particular form effectively in largely free oral interaction), it frequently becomes clear in a subsequent assessment, or even the next lesson, that the item has, in fact, still not entered learners' interlanguage (their current stage of L2 development) and is not being used correctly. Consequently, proponents of task-based approaches to language learning prefer to foreground meaningful interaction and real-time use of language before focusing learners attention on language form, in effect reversing the PPP paradigm (see Module 4).

It is understandable for language tutors to feel somewhat confused by these apparently contradictory trends. However, sound practice is likely to be one that avoids the pitfalls of extremes of methodology (eg an over-rigid sequencing of pedagogic phases or an entirely unstructured communicative 'free for all') and is mindful of the learner's need for a clear structure to his or her learning. In developing speaking at the beginner's stage, this would mean a place for repetition and drilling, and for a gradual move from short to long utterances, from simple to complex language, from scripted or didactically prepared to authentic and less salient language. Beyond the beginner's phase it would mean recognizing the need for interaction, for increasing real-time authentic language use and increasing complexification.

At intermediate and advanced levels it is often simply assumed oral work means discussion. But careful thought needs to be given to how precisely we can get learners to the point of being able to conduct meaningful conversations. Field (1999: 197-202) proposes a five-part sequencing of speaking activities. The following is a summary of this:

• Presenting a unit (eg in a textbook or a dossier): this is aimed at activating existing knowledge relevant to the topic and existing language ability; it involves closed questions and answers and maybe also repetition, response to visual aids and written texts.
• Factual responses: this involves relaying facts via oral presentations, eg narrating stories, reporting events, oral reconstruction of skeleton texts or freeze-frame video footage.
• Personalization: following the inputting of relevant information at the above two stages, this moves learners from the 'concrete to the abstract', away from roles towards more speaking for oneself. The teacher has a key role to play here, focusing learners' attention first on range, then on range and fluency, before seeking to develop range, fluency and accuracy together (cf Johnson, 1989), challenging learners' points of view, introducing unpredictable elements and using carefully selected questions to force learners' to focus on accuracy (eg specific tense usage).
• Opinion building: here unscripted oral presentations are intended to help develop learners' independence and fluency; learners need to be encouraged to use 'visual and verbal prompts' and to integrate verbal signposting or discourse markers (Johnson, 1989) to help structure their presentation (ie 'firstly, secondly', thirdly', 'on the one hand…on the other', 'I come now to my second point', etc, etc).
• Discussion: discussions amongst groups of students, ideally chaired by the teacher, allow speakers to expand on the theme of a presentation, 'thereby practising independence, fluency and spontaneity' (Field, 1999: 202). Responses from listeners challenge the speaker to clarify or to justify what he / she has said, thus creating genuine information gaps and increasing unpredictability.