The three groups of newsreaders tested (1937-1947, 1957-1967, 1977-1987) show patterns of voice quality settings that, while also bearing basic similarities one to another, identify in minor ways the decades that they reflect. The acoustic analyses support these findings. Specific patterns of formant values have been identified which can be seen to coincide with the generalized vocal settings for the groups. A dramatic change in voice dynamics has been identified that does not conform strictly to the organization of the three time period groupings.

The predominant voice quality settings are lowered larynx, open jaw, and the use of creaky phonation. The majority of newsreaders across all three time periods display these features. The other settings examined are shared in various combinations by the groups (see Table 2). The high incidence of the lowered larynx, rounded (protruded) lips, and the open jaw is seen as a voice setting configuration suitable to speakers trained to enhance vocal resonance (Lessac, 1960, Linklater, 1976). The earliest group (1937-1947) is distinguished by the highest incidence of both tongue backing and the spread-lips configuration, while in the two later groups (1957-1967 and 1977-1987) tongue fronting and rounded lips are dominant. The fronting of the tongue - and possibly open jaw - might be attributed to the increased influence of the American accent on Canadian English since the Second World War (Warkentyne & Brett, 1981). Nevertheless, the overall voice quality configurations for CBC Radio newsreaders do differ both from those identified for American English by Esling & Wong (1983), and from those for British English by Honikman (1964).

The individual vowel analyses suggest a vowel shift in progress. Although the lack of the vocalic opposition in words like cot/caught typical of Canadian English was observed in the newsreader speech, an auditory impression of a gradual, slight fronting of the low back vowel /*/ was gained from listening to the thirty newscasts chronologically. In addition, the low front vowel /*ash/ shows a significant correlation in F values among CBC newsreaders of the 1977-1987 period only. Taken together, these two are seen as a "solidifying" of /*ash/ in order to "make room" for the fronting of /*/ and to preserve the distinction between the low vowels.

Statistical tests of vowel formant values show that the speech of CBC Radio newsreaders cannot be associated with any particular class of the Vancouver Survey. The ranges in combined averaged F values do not vary greatly for the two low vowels /*ash/ and /*/. However, the differences in the averaged F1,F2 values for the high back vowel /u/ suggest auditorily perceptible differences across the two informant groups. The finding of a correlation of formant values only within CBC vowels suggests that newsreading represents a more defined standard than does the Canadian English of the Vancouver region. The formant measurements and statistical tests of /u/ showed extensive variation among both CBC Radio newsreaders and Vancouver Survey informants. The high back vowel phoneme remains undefined for Canadian English.

Both the examination of voice quality settings and analyses of vowels taken from the three test periods show a far greater uniformity in speech production during the middle period (1957-1967), just prior to adjustments in CBC Radio newscast format. Changes in editorial and management policy around the mid-1960s resulted most noticeably in great modifications to the tempo of newsreading. The news announcer was no longer required to read for fifteen minutes straight. Many breaks in the text were introduced through the inclusion of reporters' voices, and through turn-taking by two newsreaders on the major newscasts, World at Six and World Report. Sentence length has not changed considerably, but phrases have been lengthened and the pauses have been shortened (see Tables 7 and 8). A drastic reduction in the percentage of silent time within the newscast has been the result. At the same time, no reduction of articulation rate, measured in syllables per second, was apparent. These results indicate that the newsreaders are exceptionally skilled speakers.

The voice dynamic component in CBC Radio newscasts has changed over the years. The pause is no longer used so extensively to clarify sentence and story organization for the audience. An assumption is that pitch fluctuations are now used far more in order to structure of the text orally. This greater pitch movement was observed during the preparation of the thirty transcriptions, but the effect on comprehension of the newscasts (for the author) was not perceptibly different from one rate to the other. This clarity is the result of the controlled articulation rates of the newsreaders.

The CBC is an excellent subject and source for the study of radio speech. However, not all policies and procedures of the past fifty years are conveniently well-documented. The primary, ongoing concern of the Corporation has been the production of quality programs. Early recording technology was cumbersome, and those involved were not always concerned with the potential historical value of their work. For the most part, radio broadcasts dispersed into the air at the moment of presentation. Through the years, CBC departments have grown and shrunk, been decentralized and recentralized, and have undergone numerous changes in personnel and priorities. Conflicts over limited space (and time and money) have caused large volumes of stored materials to be summarily trashed. For many years preservation was arbitrary and haphazard, and as a result, gaps exist in both audio and written records. Nevertheless, huge amounts of broadcast material have been amassed in various private, corporate, and public collections across the country. These resources are still in the process of being catalogued.

The thirty transcriptions of this study constitute the first such collection of CBC newscasts. Copies of the transcriptions, complete with the names of the newsreaders, have been sent to the CBC Radio Program Archives and the CBC Reference Library in Toronto, and to the NFTSA, Ottawa to be available for use in future research.

Recorded broadcasts are readily-available resources for linguistic studies in formal and informal speech styles. Because of the findings regarding controlled articulation rates, CBC Radio newscasts might prove to be particularly suitable for psycholinguistic tests of the effects of intonation patterns and speech rate on listener perception and comprehension. The results of this study of CBC Radio newsreaders also await comparative acoustical analyses with samples from other Canadian studies. A comparison with the speech of Ottawa informants (Woods, 1979) might be particularly rewarding, considering the extensive phonological comparison between Ottawa and Vancouver that has already been done (de Wolf, 1988).

The accent of CBC Radio newsreading is General Canadian; the stronger regional dialect accents such as Ottawa Valley and Newfoundland are not usually heard on the CBC National News Service. The voices of CBC Radio newsreaders are not strictly uniform but, collectively, they represent a national standard for formal Canadian speech.

The role of the radio newsreader has changed over the years. From being one of sole dispenser of information, the newsreading task has expanded to include introducing and summarizing on-site voice reports. Changes in newscasting are likely to continue into the future, particularly as a result of developments in communications technology. Extremely compact, portable transmitting equipment and satellite relays now allow reports to be received instantly from all around the world. Since the technology exists, it is conceivable that one day newscasts could consist entirely of reporters speaking from remote locations without a newsreader to create a bridge between them. Yet, the human element in communication remains. The introduction of the assorted accents inherent in a "global news" format would result in a loss of the accustomed CBC standard. The current generation of CBC Radio listeners might not readily accept such a drastic shift in news style. The majority of older listeners still prefer to hear male, rather than female, newsreaders (CBC, 1988 Qualitative radio study). Language attitudes are slow to change and CBC Radio is highly responsive to the results of its audience surveys. News language does not introduce speech innovation but, rather, reflects the expectations of the speech community. For these reasons, the voices of CBC Radio newsreaders serve well to illustrate conservative, formal Canadian speech throughout the decades.

-30-