6 Assessing productive and interactive skills Personal refl ection

 Part 
II
Once the linguistic message has been articulated, both speakers 
and writers have opportunities to monitor and, if necessary, revise 
what they have said or written. Writers have the chance to do this 
before sharing their texts with others. Speakers engaging in interaction 
have much more limited time for their self-monitoring, but are more 
likely to receive immediate feedback (perhaps by noticing their own 
slips, or from questions or confused facial expressions from their 
conversational partners) indicating that communication has failed 
and that they need to reformulate the message. Monitoring and 
revision can occur at the local levels of spelling, pronunciation, 
punctuation and grammar. They may involve organisation of the 
overall utterance or text, levels of formality, text structure, stylistic 
elegance and the effects on the audience. In the course of composing 
a text or participating in interaction, addressers may modify or even 
abandon their original goals.
Learners are often categorised according to how they monitor and 
revise their speech. Some prioritise local accuracy. As a result, they 
tend to speak relatively hesitantly, concentrating on grammatical 
formulation and pausing frequently to plan or to correct themselves. 
Others prioritise fl uency, sacrifi cing a degree of accuracy to maintain 
the fl ow of speech. Both strategies have benefi ts and costs for the 
learner and can cause strain on the addressee. Assessment designers 
need to decide how best to take account of such choices when scoring 
performance (Fulcher, 2003).
A longstanding criticism of process models of language is that they 
tend to focus rather narrowly on how messages are transmitted 
between individuals without paying suffi cient attention to the social 
constraints and conventions that govern when it is appropriate to 
speak (or write) and the impact of these upon meaning. By focusing on 
the individual mechanisms of language processing, they downplay the 
shared nature of communication. As Luoma (2004) expressed it, they 
are concerned with 
speech
, the product of the individual, rather than 
with 
talk
as a shared social activity. He and Young (1998) argued that 
the social nature of talk raises questions about the validity of interactive 
speaking tests. Scores are awarded to the individual assessee, but talk 
is a product of shared interaction.
A further challenge for assessment designers is that the meaning of 
any sentence or speech unit is not universal, but can be radically 
modifi ed by context. An utterance such as ‘
Good evening, madam
’, 
even if spoken with identical intonation, has a very different implication 
when it is said by a hotel receptionist greeting a guest at 6 pm than 
when it is said by a mother to rebuke her teenage daughter for waking 
up late in the morning. It can be imagined that the expected response 
would also be quite different in each case. Each utterance conveys 
something about the relationship between the participants. The 

---