Looking at genres from a phonological perspective
Lucía Inés Rivas
National University of La Pampa- Faculty of Human Sciences- Foreign Languages Department
Santa Rosa, La Pampa, Argentina
Discourses have been extensively described in terms of their generic configurations: stages and phases of
genres have been characterized according to the functions they fulfil within the text as a whole, and the
choices made in the lexico-grammar and the semantics have been explored; but little has been said about
the phonological realizations of genres in oral texts. In the understanding that “the prosodic resources of
the phonological system, its intonation and rhythm, are […] central to the workings of a language […]”
and that “phonology is an essential component in the making of meaning” (Halliday & Greaves, 2008, pp.
74, 79) we set out to explore oral texts, looking for correspondences between the lexico-grammatical and
semantic patterns described and their phonological realizations. Our analysis has been based mainly on the
developments in SFL phonology and on Discourse Intonation Theory, and we have observed linguistic
features within the systems of Tonality, Tonicity and Tone, and paralinguistic ones, concerning tempo,
articulation, extension of segments, volume, pitch span, among others. Our auditory perceptions have been
validated by means of the computer software Praat. As a result, we have found some consistent tendencies
in the realizations of both textual and interpersonal meanings through particular phonological choices.
1 Background to this presentation
This presentation includes results from a research project carried out at the National University
of La Pampa, Argentina, in which we have explored oral texts in English, and the importance
of intonation choices in the configuration of oral genres. Observing oral texts from the genre
theory point of view (Martin & Rose, 2007, 2008) allows us to consider them as semantic units
in themselves, units which realize social processes with a certain communicative function.
Martin & Rose (2007, 2008) describe different genres, mentioning their constitutive stages and
phases with the communicative functions they fulfil. These authors also refer to the lexico-
grammatical and semantic choices which characterize stages, and those which may signal a
change in stage or phase. However, in this characterization little is said about the phonological
choices in oral texts.
As regards intonation systems, some studies focus on the meanings projected by intonation
choices in units larger than the tone unit. Brazil, Coulthard & Johns (1980), Brazil (1997),
O’Grady (2010, 2013, 2014), Tench (1996, 2011), Wennerstrom (2001), Wichmann (2000),
among others, refer to the textual function of intonation in segmenting texts into ‘paratones’
(‘pitch sequence’ for Brazil et al.), in a way similar to a paragraph in writing; and the role of
tone in the staging of information (Halliday & Greaves, 2008, O’Grady, 2013). Besides, the
interpersonal function of intonation has been explored with respect to the attitudes and the
communicative functions projected by speakers within a certain context. Some work by Tench
(2005) and O’Grady (2014) also relate these meanings to the constitution of texts as a whole,
from a generic configuration.
An approach towards the meaning of phonological choices from a genre perspective allows
for explanations grounded in a wider context, that of the whole text as a semantic unit, instead
of the local meaning accounts that derive from the micro context of each tone unit. In this
understanding, we set out to explore phonological choices from a generic perspective.
1.1 Corpus and method of analysis
As a corpus for our study we collected recordings from the internet from various sources. We
observed oral texts taken from different media such as the BBC, CNN, NPR, Radio Times, The
Guardian, and others; we explored interviews of different kinds, stories read aloud, phone-in
radio programmes, lectures and essays on various topics, among others.
Once we had listened to the whole text or programme, we carried out a macro analysis in
order to find segments which were suitable for generic description. We then concentrated on
those sections of the texts, transcribed them, and analysed them in terms of their generic
configuration, following SFL theory (Eggins & Slade 1997; Martin & Rose, 2007, 2008).
After that, we continued with the phonological analysis to explore whether we could find
correspondences between lexico-grammatical and semantic choices and phonological ones. We
based our analysis on SFL description of intonation systems (Halliday & Greaves 2008; Tench
1996, 2011), in combination with the Discourse Intonation systems (Brazil et al. 1980; Brazil
1997). Besides, we also observed paralinguistic features such as tempo, volume, pitch span,
placing in the voice range, clarity in the articulation, voice quality, among others (Brown,
1990). We carried out a perceptive analysis and then validated our results with the software for
speech analysis Praat (Boersma & Weenik 1992-2017).
Within SFL, Tench (1996, p.2) distinguishes between a linguistic and a paralinguistic
dimension of intonation. The linguistic one concerns choices the speaker makes within the
systems of Tonicity, choosing where to place the tonic; Tonality, the way the material is
chunked into information units; and Tone, the choice of pitch movement on the tonic. The
paralinguistic dimension concerns the messenger, “the speaker’s state of mind, their degree of
politeness and their effort to associate or dissociate” from the listener.
Discourse Intonation (DI) systems comprise Prominence, the words the speaker chooses to
highlight as a ‘sense selection’ from an existing paradigm; Key and Termination, which involve
selections on pitch level in a three element system of high, mid and low, on the first prominent
syllable in the tone unit (Key) and on the tonic syllable (Termination); and Tone, which implies
a selection in pitch movement between proclaiming –the fall and the rise-fall – referring –the
fall-rise or the rise – and the level tone. These systems, though not equivalent to the SFL
systems, can be complementary in the analysis of texts (See Pascual et al. 2010, 2013).
Prominence roughly coincides with salience, though DI emphasizes the decision on the part of
the speaker to make a particular word prominent. Though pitch level is included in SFL
phonology as a higher level of delicacy for each tone, in DI these systems are independent from
Tone, and add a particular meaning of their own.
2 Textual meanings
In the oral realization of texts, intonation has an important role in projecting textual meanings.
“Speakers’ intonation choices project an organization on their texts, which they use to guide
their hearers’ interpretative choices” (O’Grady 2013, p. 126). In this way, we started our
analysis of corpus samples focusing on the phonological realizations of the different stages and
phases that constitute genres, and the marks that may signal a change in stage or phase.
2.1 Introductions to interviews
One of the text-types we explored was what we call ‘general interest interviews’, in which the
host of a radio or television programme interviews an expert or authorized voice on a certain
topic, relevant in the news at the time of the interview. These are extended interviews (between
10 and 30 minutes long approximately) in which interviewees are asked about their opinion on
a usually controversial matter.
We noticed that the introduction to the interview, which the interviewer produces before
greeting their interviewee, was prone to generic description. These are short, usually formal
pieces of language, which although apparently unscripted, do not show the typical dysfluencies
of spontaneous speech, and thus seem to be carefully planned beforehand. They exhibit a
simple syntax, with a few clauses per sentence, and a synoptic structure which presents the
main parts of the message in a clear and concise manner. We defined these texts’
communicative function as ‘enticements’, a text with the purpose of informing the audience
about the interview to come and of seducing it to stay tuned and watch or listen to it. As regards
this genre’s composition, we recognised three recurrent stages in this part of the interviews
analysed (ten in all): Guest presentation, Link with current news, and Topic introducer; and
another stage present in seven of them which we called Addressing the audience. What follows
is a transcript of one of these texts (Text 1):
Text 1: BBC HARDtalk - Tracey Emin (http://www.youtube.com/watch?v=DH1F4Xrf0Uk)
Addressing the audience
Welcome to Margate, a traditional English seaside town which is home to the Turner Contemporary art
My guest today is Tracey Emin, who was raised in Margate and who’s become an artist of international
Link with current news
–she has an exhibition currently on in her old home town.
Her work is always deeply personal. She’s made an extraordinary journey from wild youth to pillar of the
British cultural establishment.
But just how blurred has the line been between her art and her life?
In our initial analysis of all the samples, we described a total of 47 of these stages, observing
and describing the changes from one stage to another signalled by lexico-grammatical and
semantic choices. In our phonological analysis we found a consistent tendency of speakers’
selection of high Key (in DI, high pitch on the onset syllable, first prominence in the pre tonic
segment) to signal these changes –Key selection is a relative value, the result of the comparison
with the preceding Key choice. The high pitch choice realizes the beginning of something new,
different or contrastive with what has come before. Predictably, this pattern occurred in 40 out
of the 47 cases (representing an 85%). Another important prosodic feature in the transition
from one stage to another is the declining of pitch towards the end of the finishing stage,
frequently reaching a low value on the Termination, and the presence of pause.
Figure 1: Praat Image showing pitch line at the beginning of ‘Guest Presentation’ stage.
Figure 1 above, a picture taken from Praat, shows one of those stage shifts from the sample
quoted above (Text 1), the change from Addressing the audience to Guest presentation. The
pitch window shows the previous declination with a preceding Key choice at 156.2 Hz on the
syllable ‘ART’, followed by a low Termination on ‘GAL’, and an initial Key at 242.3 Hz on
‘GUEST’. A pause of 0.27 seconds separating the stages can also be observed.
The seven occasions in which speakers chose a Key different from high at the beginning of
a stage concerned mid-Key choices. This selection implies that the section adds to what has
happened before in the expected way, without any contrast intended. Four of these choices start
the optional stage Addressing the audience. In these cases, this initial key choice can be
explained by the fact that this is the opening stage of the text, happening immediately after the
programme’s typical musical backdrop. The host addressing the audience is what is expected.
The three remaining cases correspond to the ‘Guest presentation’ stage, and represent
instances in which the order of the stages and/or the lexico-grammatical structure chosen
postpone the name of the guest to later in the utterance. In this way, the beginning of this stage
in Text 3: “My guest today is psychiatrist Dr Steve Peters”, is realized with mid Key on the
syllable ‘GUEST’, to climb up to high Key on ‘STEVE’, in a later tone unit. In this particular
case, the Guest presentation stage happens after the Link with current news and the Topic
introducer, and thus the beginning of this stage is presented as expected.
The Praat image below (Figure 2), shows the pitch line with the declination on the final
tone unit of Topic presentation, with a Termination on TAL on 136 Hz, low pitch for this
speaker; a pause of 0.43 seconds –the pitch line in the picture showing background music –
mid Key on GUEST at the beginning of Guest presentation stage and a climb up to high Key
(205 Hz) in a later tone unit in which the guest’s name is introduced.
Figure 2: Praat image showing the beginning of the ‘Guest presentation’ stage in medium pitch
and the jump up to signal the name of the guest, later in the stage.
In many of the samples in our corpus we have recognized segments of texts that can be
described as belonging to one of the genres in the family of ‘stories’. Following the same
procedure described above, we have reached similar conclusions. In a series of three interviews
from the BBC, we identified ‘observations’ produced by the interviewees. After describing the
different stages and phases and carrying out the phonological analysis, we found that in 62.5%
of the cases the change in stage or phase was signalled phonologically with high Key. The
remaining 37.5% showed mid Key, and these cases usually corresponded to stages or phases
which were very short. (Rivas & Germani, 2016).
In the case of a story read aloud from the Storynory web site, we found that 10 out of 12 stages
started with high Key, and in the case of these stories the stages were also separated by long
pauses (3.5’’ average).
Within radio interactions in two phone-in programmes from Radio 4, we described ‘recounts’
and one ‘narrative’. These stories are different from the others previously described, because
they tend to be co-constructed between the caller to the programme and the host herself. It was
interesting to see that when the host interrupted the caller because she wanted to advance the
recount or narrative, her interruption caused certain disruption, as it usually started with a pitch
higher than the one being used by the caller, louder volume and faster tempo, to indicate a
change in stage. But when the host’s intention was that the caller expanded on what s/he was
saying without changing stages in the generic contribution, these interruptions were done in
medium pitch, and volume and tempo similar to the caller’s speech.
3 Interpersonal meanings
Intonation realizes interpersonal meanings. The analysis of these meanings depends very much
on the particular genre and the contextual features in each case, but there are some regularities
that can be mentioned.
3.1 Introduction to interviews
In these short texts, the interpersonal function comes at the forefront in the stage ‘Guest
presentation’, as the host introduces their guest in a way that seduces their audiences to stay
tuned. Given this function, most of the appraisal choices speakers make are related to
judgement and engagement, and they show positive values, aimed at generating alignment of
views in the audience. Phonologically, most of these short utterances are given falling
intonation, a ‘telling’ tone which suggests the speaker is presenting a piece of news,
information meant to enlarge the area of common ground shared with the audience. The choices
in Tonality segment texts into very short units, emphasizing the information presented. Besides,
particular phonological choices are used on the realization of appraising words, which are
focused on by means of jumps up in pitch, which result in high Key choices in the interior of
stages and phases. This ‘interpersonal’ use of Key suggests a particular unexpected sense
selection within an existing paradigm of options, and highlights the word’s semantic load. The
transcript below shows the choices mentioned above.
Figure 3. Transcript of one ‘introduction to interview’ with intonation marks.
In the transcript above, tone units are separated with double bars (‖). Prominent syllables
are transcribed in block capitals, tonic syllables underlined; high and low pitch choices are
signalled with small capitals H and L respectively before the salient syllable, mid pitch remains
unmarked. Tone is shown before the tonic syllable with a tonetic mark: for fall, for rise,
for fall-rise, for rise fall and > for the mid-level. As can be observed, high pitch highlights
the interpersonal semantic load in the words “fifty”, “rocking”, “also” and the agent and process
in the projecting clause “he believes”.
This presentation shows just a sample of the work done as regards the phonological realization
of different genres. The limitation of space does not allow for a more exhaustive presentation.
These explorations have had the shape of ‘case-studies’, given the limited number of texts
collected and analysed for each type, under our working conditions and resources.
Nevertheless, we have reached preliminary conclusions in the form of tendencies which need
to be further tested with more research to come. Our results, though limited with respect to
their representativeness, do show consistent tendencies in the role phonology plays in both
signalling and realizing textual and interpersonal meanings. More work is necessary to confirm
A genre approach to the meanings projected by the phonology allows us to demonstrate
what Halliday and Greaves (2008) claim, that “the prosodic resources of the phonological
system, its intonation and rhythm, are […] central to the workings of a language […]” and that
“phonology is an essential component in the making of meaning” (2008, pp. 74, 79).
Boersma, P. & Weenink, D. (1992-2014): Praat. Doing phonetics by computer (Version 5.4.02) [Software].
Disponible en http://www.praat.org
Brazil, D., Coulthard, M. and Johns, C. (1980). Discourse intonation and language teaching. London:
Brazil, D. (1997). The communicative value of intonation in English. Cambridge: C.U.P.
Brown, G. (1990). Listening to spoken English. (2nd ed.) London: Pearson Education.
Halliday, M. & Greaves, W. (2008). Intonation in the grammar of English. London: Equinox.
Martin, J.R. & White, Peter R. (2005): The language of evaluation. Appraisal in English. Hampshire, Palgrave
Martin, J. & Rose, D. (2008). Genre relations. Mapping culture. London: Equinox.
Pascual, G., Germani, M. & Rivas, L. (2010). La unidad tonal. Una mirada comparativa de la entonación en la
conversación en inglés. In Castel, V. & Cubo de Severino, L. La renovación de la palabra en el bicentenario
de la Argentina. Los colores de la mirada lingüística (pp. 987-993). Mendoza: Editorial FFyL, UNCuyo.
Pascual, G., Rivas, L. & Germani, M. (2013). El movimiento tonal: una mirada teórica comparativa de la
relación entre entonación y significado en interacciones en inglés. In Labastía, L. Cuestiones de Fonética,
Fonología y Oralidad (pp. 33-48). Mendoza: Editorial FFyL, UNCuyo.
O’Grady, G. (2010). A grammar of spoken discourse. The intonation of increments. London: Continuum.
O’Grady, G. (2013). Choices in Tony’s talk: Phonological paragraphing, information unit nexuses and the
presentation of tone units. In O’Grady, G., Barlett, T. & Fontaine, L. (eds). Choice in language. Applications
in text analysis (pp. 125-157). London: Equinox.
O’Grady, G. (2014). The use of Key in projecting face-threatening acts in televised political debate. Text and
Talk, 34 (6). (pp. 685-711).
Rivas, L. & Germani, M. (2016). Analysing correlations between generic patterns and prosodic realizations in
interviews in English. International Journal of Language Studies, 10 (2). (pp. 103-126).
Tench, P. (1990). The roles of intonation in English discourse. Frankfurt am Main: Peter Lang.
Tench, P. (1996). The intonation systems of English. London: Casell.
Tench, P. (2005). Intonation and categories of meaning. Cardiff University, Wales, UK. Available at
Tench, P. (2011). Transcribing the sound of English. Cambridge: CUP.
Wennerstrom, A. (2001). The music of everyday speech. Prosody and discourse intonation. Oxford: OUP.
Wichmann, A. (2000). Intonation in text and discourse. Beginnings, middles and ends. London: Longman.