Meta-pragmatic discourse is defined as the feedback which speakers provide to their own as well as to others' language behavior. This meta-pragmatic discourse examines, reviews, and comments on the correctness and exactness of words and expressions. Therefore, meta-pragmatic discourse is an expression of the ability to criticize human thinking and behavior – and as such, is related to literacy.

Metapragmatic comments identified in the data were coded in three fields: a) talk about discourse management, b) meta-linguistic talk (Blum-Kulka, 1997; Aukrust, 2001) and c) reported speech. (Aukrust, 2001). The results show that the children from both groups and in both situations prefer to talk about discourse management as their first choice and meta-linguistic talk as their second choice. Qualitative analyses reveal that meta-pragmatic talk deals with three subjects; the mechanisms of conversation, dialogicity (Hamo & Blum-Kulka, in print) and literacy. The research suggests directions on how to use meta-pragmatic discourse to encourage social and academic learning.

The goal of the research was to map children's meta-pragmatic discourse and to explore how yet in class wide learning. An additional aim was to investigate developmental aspects by comparing different age groups with regards to their meta-pragmatic behavior.

Research designThe research presented is part of a three-year longitudinal study on the development of discourse skills of children in two age groups: 4-7 and 9-11 year olds (Blum-Kulka, 2004). Each age group comprised 20 children, a total of 40 children. The data were collected in two situations - natural peer talk and half-structured interviews with adults. Meta-pragmatic comments identified in the data were coded in three fields: (a) talk about discourse management, (b) meta-linguistic talk (Blum-Kulka, 1997; Aukrust, 2001) and (c) reported speech (Aukrust, 2001). Data were analyzed, both quantitatively and qualitatively. Data presented is the average of meta-pragmatic remarks, in each of the three fields, made in a unit of 20 minutes per child.

The main findings of the research showed that the children from both age groups and in both situations (peer talk and interviews with adults) prefer to talk about discourse management as their first choice; meta-linguistic talk as their second choice; and reported speech as their last choice.

Qualitative analyses reveal that meta-pragmatic talk deals with three dimensions: (a) the mechanisms of conversation, (b) dialogicity (Hamo & Blum-Kulka, in print) (c) and literacy. As part of the first dimension, i.e. mechanisms of conversation, findings showed differences in the pattern of behavior of the younger children as opposed to the older ones. While the younger children explicitly requested and negotiated their turn-taking while conversing during a pretend-play, the older children were less demanding as to their turn and were more attentive to the needs of their peers.

With regard to the second dimension of dialogicity, it was found that both groups, in both situations, intended to the conversational rules including quantity, quality, manner and relevance (Grice, 1975). Using comments on these four conversational rules children were engaged in childhood norms as well as norms of the grownups. With respect to the third dimension of literacy, it was found that children, in all age groups, use extended discourse which is remote from the here-and-now. They also use cognitive verbs in both situations, but more so in the interviews than in the peer-talk. It was also found that children are constantly engaged with words, are preoccupied with the reading and writing skills and display a wide perception of what literacy is all about. In doing so, children deal both with school literacy as well as with life literacy.

In conclusion, it can be seen that the findings of this research corroborates previous research (Hoyle & Adger, 1998) which showed that meta-pragmatic skills are already evident at a very young age, even as young as 3 years of age. The findings of the current research show that this meta-pragmatic knowledge continues to increase and develop further on with the age of 9-11. The fact that children make meta-pragmatic remarks shows us that discourse management is not automatic yet and requires constant engagement and processing. From the above-mentioned results it can be attended to use that meta-pragmatic discourse is meaningful for the advancement of social and discourse learning, among children who intended to use their mechanism of talk, dialogicity and literacy skills in spoken language. Since meta-pragmatic knowledge is strongly identified with learning processes it is of utter importance for teachers (and teacher-trainers) to be aware of this and to make sure that a variety of educational activities, which require the use of meta-pragmatic skills, are proposed to children in order to promote their social and academic learning.