Assessing the Benefits of 4/3/2 and 3/3/3 Speaking Activities

Fluency development activities are often neglected in classrooms. Perhaps this is because some teachers feel they need to actively teach the students new material, as opposed to practicing or teaching fluent use of language already acquired. I chose to give four people two fluency activities in the form of a 4/3/2 task and 3/3/3 task. These tasks work by giving the students a topic to talk about, a few minutes to think about the topic and then speaking about the topic for a set period of time. In the case of the 4/3/2 task, students discuss the same topic for four minutes, then three minutes and lastly two minutes. In the 3/3/3/ task, they discuss the same topic with the same three minute time restriction each time (Nation, 1989: 378).

Both activities are aimed at improving fluency, or the processing of real-time language use (Nation & Newton, 2009:151). In this essay I will be mostly looking at fluency in terms of speed and flow. However, to draw on Paul Nation’s definition (1989: 377), I will also look at the amount of control the learner has over the language used, and the way this language interacts with the meaning. Fluency is a skill in itself and although accuracy can inhibit how fluent someone is perceived, it is both possible to “speak correctly but not very fluently” (Lennon 1990 :390) and to speak fluently while making grammatical errors.

The first time the learners respond first with free production, as it is in essence an unprepared speech (Nation, 1976: 8), and then re-do it subsequent times. This works as both practice and as a chance to re-work the material, by ironing out mistakes and excessive material. The tasks themselves facilitate fluency development through meaning-focused output, which along with meaning-focused input and language-focused learning, make up the central aspects of second language acquisition. Collectively, these aspects are known as the four strands of language learning (Nation & Newton, 2009: 1-14).

I gave both tasks to two non-native speakers (NNS1 and NSS2) and two native speakers (NS1 and NS2). All subjects consented to the material gathered being used in this essay.

Each subject discussed what they do in a typical week for the 4/3/2 activity, and what they would like to do in the future for the 3/3/3 activity. Approximately one hour separated the activities, which I thought would limit fatigue. Usually the subjects would change partners each time they did the task, but unfortunately this was not possible and the subjects talked to me each time.

How the output evolved for 4/3/2 and 3/3/3

As a result of repeated tellings, the subjects’ output changed significantly after both the 4/3/2 and the 3/3/3 activity. The NNS subjects were hindered in the initial dialogues for both activities because of excessive hesitation items, pauses and unconnected speech. However, by the third time, they made greater use of formulaic sequences, such as collocations and connected speech, as well as significantly less hesitation items.

In terms of total output, all learners increased their number of words said, while reducing the number of hesitation items, such as ‘you know’, ‘ah’, ‘like’ and ‘um’ in the 3/3/3 activity. They spoke with more confidence after the initial dialogue, and seemed to be more focused on the language they were using, as the story itself was already established.

Generally, despite decreasing speaking time, the learners managed to cover the same content in the 4/3/2 activity. In part, this was because there were less recasts, repetition and hesitation items. One example of this comes from NNS1 who said the following when discussing what she does during the week:

Four minute attempt: …Or + /I’m now ah\ reta-retail shop+ /Clothing shop\ (07 seconds)…

Three minute attempt: …I + /I’m working at\ clothing shop (04 seconds)…

Two minute attempt: …/I work at clothing shop\ (02 seconds)…

This highlights how the language become more fluid, with less pauses, false starts and interruptions. Doing so helped the learner to communicate the same information in 2 seconds as opposed to 7 seconds, and to do so with greater grammatical accuracy and fluency.

The NSs struggled somewhat with the 4/3/2 activity, as I did not mention that the time would be getting shorter with each attempt until after the first four minute dialogue. Both students managed to cut out many of the hesitation items, and generally spoke at a faster pace. As they both covered quite a bit of content and did not struggle with the actual language, the 2 minute telling ended up with both students speaking unnaturally fast. The NNSs covered less content on their first telling, and although subsequent tellings skipped some details, they did not sound rushed. This is largely to do with the fact that the NNSs had more room for improvement than the NSs.

The use of formulaic speech helped the NSs in their speech significantly. They made much use of connected speech from the beginning, but also used it more when the time pressure was increased.

An example of NS1’s speech is as follows:

…/In the weekend\ I usually just /go into town\ and /get on the piss\ with the boys…

…/In the weekend\ /I hit the town\ and meet up with /my mates\…

…/In the weekend\ /I drink in town\…

Three of the four subjects used the formulaic sequence ‘in the weekend’ multiple times, while NS2 did not use it because she differentiated between what she does on Saturday and Sunday. The use of such phrases requires less labour, and is therefore faster and can be said more fluently (Wood, 2009: 41-43).

Learner Feedback

Generally, the learners found the 4-3-2 task to be slightly restrictive and there was a comment that maybe it would be easier if the task became shorter at a slower a pace. The subject suggested a “4-3:30-3” task. However, both NS1 and NNS2 found the task to be good as it forced them to produce clearer, more direct sentences while thinking on their feet. The feedback was that in both activities, by the third attempt they were more confidant in their speech and that their fluency would improve as a result

Learners generally found the 3-3-3 task easier because there was less pressure from the clock. The general consensus seemed be that if they only had to cover as much content in the second and third attempts as they did in the first, then there was less to worry about. This feeling helped the learners to relax on the second and third attempts, and although they were nervous the first time, they felt more confident in the subsequent tellings.


Both activities achieved the goal of improving fluency, but had slightly different results in terms of how that fluency was realised. In both of the activities, the first time through the subjects generally had an emphasis on content and getting the information down. The structure of this information evolved with subsequent re-tellings in all cases.

In the 3/3/3 activity, on subsequent tellings, the students seemed confident that they could get the information out and focused more on structure, correct use of language and increased lexical diversity. The way the information was articulated became clearer and easier to follow, which gave more room to incorporate and experiment more interesting phrases.

In the 4/3/2 activity, the learners developed speed, and some language deemed irrelevant or unnecessary was dropped by all of the learners. NNS2 in particular used a lot of recasts and repetition of words in her earlier attempts, but managed to drop these considerably by her last speech. The below quotes are from NNS2’s four minute and two minute speeches.

Four minute attempt:

…/There are many\ volunteering staff and different + /kind of ah\ different kind of ah + age group age people and also different ethnic people, different country people. + /It’s really different from me\ + And + it’s interesting to /talk to people\ how different each other. Like + about differences… (29 seconds)

Two minute attempt:

…/There are many volunteer staff\ who are different age /and different country people\ + It’s interesting to ah /talk to people\ like that… (10 seconds)

The difference between the two attempts demonstrates how a combination of time pressure, repetition and focusing on the message helped to facilitate the development of a higher than usual performance (Nation & Newton, 2009: 154-155). The learner also made use of the formulaic sequence ‘talk to people’ on all three occasions, which indicates that she remembers this as a set item, and can probably use it with as much ease as she can a single word. By the third attempt the learner has decided on which phrases and words to use, which means she is repeating words, especially the word ‘different’ less. She has also stopped backtracking to fix words she has perceived as mistakes.

To a certain degree, the students must make a tradeoff between speed and accuracy, especially when they are against the clock. This may be influenced by whether the students are naturally introverted or extroverted (Dewaele & Furnham, 1999: 536). Pressure from the clock in the 4/3/2 activity increased the speed of the NS students, but also lead them to better structure and content, with less unnecessary information. For the NNS students, the time restraint led to faster, less accurate speech, although it also led to increased use of previously memorised phrases and phrases from previous tellings of the story.

As a study, a bigger sample base would have provided results that one could generalise about more. In analysing the speech, I looked at total number of utterances as one indicator of improvement, but I counted hesitation marks as distinct items. I chose to do this because I considered that they are used by native speakers and non-native speakers alike as distinct utterances. This makes it easier to see how fast the learner was speaking, but it does not indicate the quality of the language. Using this system on a bigger scale could prove problematic in that although hesitation items are a part of speech, too many can decrease the appearance of fluency. An example of this is from NNS1’s first 4/3/2 speech, where she uses 23 items in 0:17 seconds, but it does not sound fluent.

…Ah+ I go out and catch up my friends and + ah + talk + like + /you know lately\ what’s + like /kind of like\ normal topic…

By her third attempt, she was using a lot less time and words to articulate this, which made her sound a lot more fluent. It was not number of words that did so though, it was that she had a good pace and did not unnecessarily use words that threw her off track.

Another problem with choosing word items for inclusion is that ‘the’ or even ‘ah’ was counted as one item, as was multi-syllable words. Producing more complex words could be an indicator of improvement, but it would not be noticed as such with the current system.

Overall, the process made by the subjects was impressive. It is beyond the scope of this essay to analyse the data in depth, but the activities proved that fluency is something that can be taught and practiced in the classroom. I had not done either of these activities prior to this, so I found it very interesting to see the benefits of repetition and time pressure. In particular, I found that the subjects’ desired meaning became more focused and easier to follow, while grammatical inaccuracies and hesitation items were significantly reduced. In part, this was facilitated by the increased use of formulaic sequences, which was an unexpected and pleasant surprise.


Dewaele, J. M. & Furnham, A. (1999). Extraversion: the unloved variable in applied linguistics research. Language Learning, 49, 509-544.

Lennon, P. (1990). Investigating fluency in EFL : A quantitative approach. Language Learning , 40/3 :387-417.

Nation, I.S.P. (1976) Creating and adapting language teaching techniques. RELC Journal, 7, 2, ,: 1-15.

Nation, I.S.P. (1989). Improving Speaking Fluency. System, Vol. 17, No. 3, pp. 377-384. Pergamon Press: Great Britain.

Nation, I.S.P. & J. Newton (2009). Teaching ESL/EFL listening and speaking. New York: Routledge.

Wood, D. (2009). Effects of focused instruction of formulaic sequences on fluent expression in second language narratives: a case study. Canadian Journal of Applied Linguistics, 10, 39-56.