In this paper we analyze variable presence of the complementizer that, like in the sentences I think that I love you/ I think Ø I love you, in a large archive by Sali A Tagliamonte. We seek to analyze the factors that favor or disfavor the variant which is the complementizer “that”. For this study the two variants are that or zero as illustrated respectively in the first and second sentence of the example. Each token was coded with five factor groups: the verbs preceding that or zero (know, say, think and other verbs as factors), the tense of those verbs (past and present as factors), the gender (Male or Female as factor groups) the age of the speaker (young, mature or old as factors) and the education of the speaker (more educated or less educated as factors). To run the analysis of the samples, the linguistic program GOLDVARB was used. The results of the best runs from the binomial up and down are presented in table 1 and in table 2 the same results are now organized into factor groups, factors, weights and ranges.
Our analysis reveals that the most dominant factor group influencing the zero complementizer is the verb before the complementizer (range 55). In addition the verb think (0.773) appears to highly favor the use of zero complementizer (Ø ) while say (0.397) disfavor Ø , know (0.282) and other (0.225) highly disfavor the use of zero complementizer.
The second most favorising factor for the zero complementizer is the tense of the verbs before the complementizer (range 24). In fact the present tense (0.573) happens to slightly favor the zero complementizer while the past tense (0.333) highly disfavors it.
With a range of 16, the age of the speakers is the third factor in favor of the use of zero complementizer. In fact, young people happen to highly favor it while old people highly disfavor the use of zero complementizer and mature speakers are neutral.
Gender is the fourth factor which favorizes the use of the zero