As the frequencies in Table 3 indicate, PhilE begins and ends the three-decade period under examination with a lower frequency of progressives (2417 → 2647) than BrE (3112 → 3340) and AmE (3090 → 3376). All three varieties evidence a mild increase, one almost identical in PhilE (+9.5%) to its colonial parent AmE (+9.3%), but slightly lower in BrE (+7.3%).3
3. Frequencies for BrE and AmE are based on Table A6.3 for AmE and Figure 6.2 for BrE (Nicholas Smith kindly supplied me with precise frequencies for the latter). Frequencies for General prose were omitted, in order to achieve generic matching for the three varieties.
Because the genres are of different sizes it was necessary to calculate averages for the three
Free ebooks ==> www.ebook777.comRecent diachronic change in the progressive in Philippine English 279 Table 3. Genre distribution of progressives across the varieties (frequencies pmw; raw frequencies in brackets)*
BrE AmE PhilE
LOB FLOB % diff Brown Frown % diff Phil-
Brown ICE-PHI % diff Press 3251
(578) 3835
(684) +17.9%** 3210 (573) 3589
(635) +11.8% 2619 (309) 3544
(241) +35.3%***
Learned 846 (136) 966
(155) +14.2% 1222 (198) 960
(154) –21.4%* 735 (61) 648
(57) –11.8%
Fiction 5239 (1344) 5218
(1343) –0.4% 4839 (1229) 5580
(1413) +15.3%*** 3898 (651) 3749
(180) –3.8%
AVG 3112 3340 +7.3% 3090 3376 +9.3% 2417 2647 +9.5%
* Asterisks indicate levels of significance as determined by the log-likelihood test: * = ‘significant at p < 0.05 (log-likelihood > 3.84)’; ** = ‘significant at p < 0.01 (log-likelihood > 6.63)’; *** = ‘significant at p < 0.001 (log-likelihood > 10.83)’.
Table 3 indicates that the rise of the progressive has not been consistent across the three genre categories. Learned writing has the lowest frequencies in all three variet- ies, perhaps not surprisingly in view of its reputation as an ‘uptight’, conformist genre (Hundt & Mair 1999), and more specifically its tendency to be concerned more with general truths rather than events of limited duration. The 1990s frequency for PhilE (648 tokens pmw) is considerably lower than that for BrE (966) and AmE (960). The distaste in PhilE for the progressive in this genre is furthermore compounded by a declining frequency (–11.8%), one that aligns it more closely with AmE than BrE.
The striking underuse of the progressive in PhilE learned writing may reflect uncer- tainty about the stylistic suitability of an apparently colloquial feature in this conserva- tive genre (compare the relative underuse in learned PhilE writing of relativiser that reported in Collins, Yao & Borlongan (2014), and the absence of contracted progres- sives in learned PhilE writing reported below in Section 10).
At the other extreme is fiction, the genre that has the largest 1990s frequencies in all three varieties – albeit only marginally higher than that for the press category in PhilE – undoubtedly a by-product of its receptiveness towards both the past progres- sive (in narrative) and towards the present progressive (in dialogue). Here it is AmE that has the highest 1990s frequency, and whose rise of +15.3% sets it apart from BrE and PhilE, both of whose frequencies remain almost static over the three decades.
The most striking result for fiction is the relative smallness of the PhilE frequencies,
frequencies in the bottom row of the table, rather than the total of the raw frequencies for the three genres normalised to tokens pmw.
www.ebook777.com
Free ebooks ==> www.ebook777.com
280 Peter Collins
compared to those for AmE and BrE (notice that PhilE fiction frequencies for another colloquial feature, quasi-modals, also were found to lag behind those for BrE and AmE in the study reported by Collins & Yao (2013). A possible explanation is that at this stage in the development of Philippine literature writers are yet to attain the same level of confidence in the techniques for speech representation that their British and Ameri- can counterparts have achieved.
The only category in which the progressive has enjoyed a rise in all three varieties is press (identified by Hundt & Mair 1999 as the most ‘agile’, innovative genre). PhilE has a comparable 1990s press frequency to BrE and AmE, but at the same time a far stronger rise (+35.3%) than the supervarieties (a finding similar to that reported by Collins, Yao & Borlongan (2014) for relative that in PhilE). Why should the Philip- pine results for press be so different from those for fiction and learned writing? One possible explanation for the PhilE Press results is that made in Collins, Borlongan &
Yao (2014), that in the Philippines there is a tendency for press reportage and edi- torials to be dominantly inward-looking, concerned with intra-national issues, with the attendant likelihood that journalists and editors will feel relatively and increas- ingly confident about using informal, colloquial, grammatical features here. In this respect press certainly differs from learned writing, with its more outward-looking orientation, usually or at least often directed at an international readership. Fiction is perhaps in-between, directed not only at a local readership but sufficiently at an inter- national audience to make writers cautious in their use of overtly colloquial grammati- cal features. The disregard for the stylistic differentiation between spoken and written conventions that Gonzales (1997) alleges to characterise Philippine writing may then be genre-specific: a genuine tendency in press, but less valid for learned writing and fiction.
The present study is based exclusively on written data. It is however relevant to mention the results for speech available from some previous studies, which suggest strongly that the progressive is a colloquial feature, and in turn that colloquialisation is a factor in its increasing frequency in writing. In order to compare how far develop- ments in the written Brown family corpora are reflected in spoken English registers.
Leech et al. (2009: 125) conclude, after examining the genre distribution of the pro- gressive in ICE-GB, that “registers that are speech-based (or speech-like) and rela- tively informal in character show a much higher incidence of progressives than typical printed genres”. The top four ICE-GB categories (telephone calls, social letters, face- to-face conversations, and broadcast discussions) are all very interactive and, with the sole exception of social letters, conducted in ‘real time’. These findings confirm those of other studies showing a significantly higher frequency of occurrence in speech than in writing (Allen 1966; Quirk et al. 1985: 198; Biber et al. 1999: 462; Mindt 2000: 248;
Rửmer 2005; Collins 2008: 230). Leech et al. (2009: 126) also explore diachronic trends in BrE speech using limited corpus evidence from DCPSE (Diachronic Corpus of
Free ebooks ==> www.ebook777.comRecent diachronic change in the progressive in Philippine English 281 Present-Day Spoken English), noting a sharp rise of over 45%, considerably larger than the 9.6% that they record for LOB/FLOB overall.
The corpus resources necessary for a diachronic examination of progressives in PhilE speech do not exist. Nevertheless it is instructive to compare the PhilE conver- sation (category S1A) results in Collins’s (2008) ICE-based study of the progressive with the findings of the present study. Collins (2008) finds the progressive to be more frequent in 1990s PhilE conversation (8250 tokens pmw) than it is in both BrE (7650) and AmE (7933), and clearly a colloquial feature in all three varieties. However in view of the relatively modest frequencies for PhilE writing noted above, it may be inferred that in PhilE the spread of the progressive has been driven less by colloquialisation than in the supervarieties.4 The role of colloquialisation is discussed further in Section 10 below.