Probability Distribution of Represented Sources in Conversations of Adults and Children

ABSTRACT

Based upon a taxonomy of represented sources, this paper investigates the quantitative features of two groups – one comprised of adults and the other of seven-year-old children – that employ a variety of sources in everyday representation. Results indicate that: 1) the overall probability distributions of represented sources used by both groups fit well to the modified right truncated Zipf-Alekseev distribution; 2) the R² value of the adult group is lower than that of the child group, largely due to the prevalence of non-present non-specified human references by the adults; 3) the values of fitting parameters a and n differ significantly between the two groups; 4) while representing from sources is an extralinguistic phenomenon, it nevertheless reflects the similar quality of language (i.e. a human-driven complex adaptive system), and it also offers a better understanding of the quantitative features of this phenomenon. In summary, this study presents the results of several preliminary attempts to study a specific type of extralinguistic phenomenon from a quantitative perspective.

Acknowledgments

Sincere thanks to the anonymous reviewers of JQL for their thoughtful comments on earlier drafts of this manuscript. I also am deeply grateful to Prof. Haitao Liu for his advice and invaluable assistance, which contributed greatly to my research.

Disclosure statement

No potential conflict of interest was reported by the author.

Supplementary material

Supplemental data for this article can be accessed here.

Notes

1. In this paper, we prefer to use the term ‘representation’ instead of ‘reporting’ because it stresses the active involvement of the speaker engaged in the act of representing or constructing another discourse (Vandelanotte, 2009, p. 10).

2. When speakers or writers represent their own discourse they treat themselves as others at another time and/or in another place.

3. The represented speech can alternatively be marked in represented dialogue by, for example, prosodic features and turn-taking structures. Both source and representing verb are not necessarily explicit in the representation, if we take zero quotatives (Mathis & Yule, 1994) as an example. We therefore restrict this study to represented discourse with dialogue introducers.

4. Except for represented source, the other three dimensions of representation are message, signal and attitude (Thompson, 1996).

5. According to the New Century Chinese English Dictionary, ‘rénjiā’ can also be used to refer to a specified person, thereby functioning like the pronouns ‘he’, ‘she’, ‘they’, etc.

6. Both standard deviation and variance can be used to measure how far the numbers spread out from the mean. In this study we use standard deviation to clarify more sensitive variability because most of these parameters have values less than 1.

Additional information

Funding

This work was supported by the Zhejiang Provincial Planning Project of Philosophy and Social Science, China [grant number 18NDJC140YB].