The Effect of the Broadcast Medium on the Language of Radio and Television Sports Commentary Genres: The Rugby Union Lineout

Abstract

This study provides provisional answers to the question, What influence does the broadcast medium have on the speech of sports commentators? It answers the question through a comparison of the influence of two media on the way in which rugby football commentaries of lineouts are given. Previous linguistic research has examined sports commentary speech largely independent of the effects of the medium in which commentary is broadcast. However it is clear that sports commentaries serve somewhat different purposes in the case that someone can see what is happening and where they cannot. We show that the differences are manifested, for instance, in whether an episode in the game is given commentary at all, the amount of commentary the episode receives, commentators' priorities recounting the various subepisodes of the game they are viewing, and the weight given to color versus play-by-play commentary. In all these cases, radio commentators provide more detailed commentary. It may be, therefore, that as far as the induction into the oral traditions of commentary are concerned, radio is a more demanding apprenticeship and that recruiting television commentators from among radio commentators will lead to a more fluent television commentary.

Introduction

Sports Commentary

Sports commentary as a spoken genre has evolved in concert with electronic media and thus has a history which can be traced.¹ Information from the early days of wireless commentary can be gained from those who participated in the evolution of their varieties of commentary (Tasker, 1985) and from early recordings (Kuiper, 1991). One significant set of anecdotes of the former concerns the interplay between telegraphy and wireless commentary. For a period in the 1930s before shortwave radio was widely available and reliable, in both Australia and the United States (Smethers & Jolliffe, 1992) commentaries were conducted on the basis of telegraphed information of what had happened in a game elsewhere. In the case of tests for the Ashes Trophy in international cricket, telegrams were sent from England to Australia, where the Australian commentators would provide wireless commentary on the basis of the information they had just received telegraphically. This shows that the available electronic media are employed wherever possible and that, by this time, commentary traditions were already well enough established for commentators to simulate live coverage on the basis of slight but critical information. Such traditions, once established, are passed on to subsequent generations of sports commentators.

With the advent of television, a new broadcast medium allowed listeners to also see for themselves what was happening or had happened. Yet the fact that, at cricket matches and horse races, spectators, despite being present at the event, routinely follow the radio commentary, if there is one, shows how important commentary is as an interpretation of the events that spectators are witnessing (McGuire, 2002). Further media innovation in the form of television replays allows commentators and, in turn, spectators to be wise after the event.

Currently, new forms of broadcast media are having further effects. For example, Eurosport broadcasts televised sports in many countries. One camera team follows the sport to provide the video, to which the audio commentary is added by native-language-speaking commentators in the countries where the commentary is broadcast. Commentary is also podcast.

Of the electronic media, all other things being equal, radio makes the greatest demands on commentators since silence is not an option, whereas periods of silence are an option for television. For this reason, radio commentary of longer sports fixtures may have pairs of commentators: a play-by-play commentator and a color commentator (Ferguson, 1983; Pawley, 1991). Paired commentary of this kind is also used in sports where there are periods of inaction, such as ice hockey, cricket, and American football, since there are interstices between action to be filled and conversations are more interesting than monologues for this purpose.

This paper will first outline some of the ways in which the broadcast medium effects sports commentary, before addressing the question, Which linguistic differences in live sports commentators may be attributed to the medium in which the commentary is broadcast.

The Commentator

Since we will be looking at the performance of sports commentators, it is important to understand how one becomes a sports commentator. Sports commentating is a profession with comparatively few members. It is not an occupation for which one can study formally. So how does one become a sports commentator? The answer is slowly. While children can be good at mimicking commentary genres (Hoyle, 1991; Hoyle, 1998), fluent, mature commentators often take about 20 years to reach that state. The major reason is that the required enculturation into the sport takes time. One should know the sport's history and all the relevant details of current players, including their nicknames (Kennedy & Zamuner, 2006); should preferably have been a known player (Justin Robert & Cummins, 2009); and must become proficient as a performer (Baker, 2007; Kuiper, 1996). For the most part, one should also be male. Female play-by-play commentators of male sports are rare (Jenkins, 1991). The major reason for this situation is prejudice (Hardin et al., 2009; Toro, 2005).

Locality Factors

Since we will be looking at the speech of sports commentators in one particular sport in one particular country, it is important to emphasize that there are few worthwhile generalizations to make about the language of sports since locality factors play a large part in sports language (Leitner, 1983; Leitner & Hesselmann, 1996; Reaser 2003). Genre studies (Askehave & Swales, 2001; Bakhtin, 1986; Martin, 2001) allow this point to be made since genre studies make local connections between contextual factors and text-type factors.

Any sport, however global, has local sociocultural manifestations. Sports commentary text types evolve locally with little boundary crossing from one sport to another or, indeed, from one country to another. Very few sports commentators move countries, although there are a few exceptions. For example, a number of horse race callers from Australia operate in Hong Kong and England. While international sports stars may sit in the comments seat, play-by-play is provided by a local commentator.

One of the reasons why there is little crossover from one sport to another is that the enculturation and skill levels required for mature, fluent commentary take so long to acquire that this is difficult to do for more than one sport. Once one sounds like a basketball commentator, one is unlikely to make an acceptable cricket commentator.

Commentary Speech

As far as the language of sports commentators is concerned, it may be noted that all sports and their commentary traditions have their own technical terms, the meanings of most of which is opaque to anyone not familiar with the sport. Field positions in cricket are a good example. The field on which a cricket game is played has a 22-yard section of turf called the pitch in its center, running along the center line of an oval field with a marked boundary. One side of this pitch is the off side and the other is the on, or leg, side. Which side is which depends on the batsman who is on strike on the pitch. Since the batsman stands at a right angle to the pitch, the on, or leg, side is that half of the field which is nearest the legs of the batsman, whereas the off side is the other side of the field. So if a batsman is left handed, as opposed to right handed, the off and leg sides of the field are reversed. When a right-handed batsman is on strike at the other end of the pitch, again the sides of the field are reversed. If this explanation is confusing, the reason is that in the absence of local knowledge, such an introduction to the culture of a sport is likely to be opaque. Baseball fields, by contrast, do not undergo these changes of positional nomenclature when the handedness of a batter changes. Sports vocabulary is thus local in that it relates to a particular game and its cultural assumptions.

The specific episode central to the discussion that follows is the rugby union lineout. In ice hockey the equivalent episode is a face-off, and players face off (Kuiper & Haggo, 1985; Kuiper, 2009). Such episodes often consist of subepisodes. For example, in cricket (Pawley, 1991) a central event occurs when a bowler bowls a ball to a batsman. For this event there is a sequence of subevents, such as the bowler's approach, the location where the ball hits the pitch, what the ball does while in the air, how the batsman responds, whether the batsman hits the ball, how it is hit, where it goes, and what the consequences are. There are specialized linguistic formulae (Lord, 1960) for all these subevents.

The cultural loading of specialized formulae exist in the wider culture of the sport since contexts have a Chinese box character (Beard, 1998). Cultural loading may be seen in formulae that involve metaphors such as those that originate in the domain of war (Chapanga, 2004). (After more than 2,000 years of sports history, contemporary gladiators are, however, not as likely to die in the arena.) Sexism and racism play a significant role in lexical selection in sports commentary in the United States, showing how general cultural proclivities may be reflected locally (Bruce, 2004;Buffington & Fraley, 2008; Denham, Billings, & Halone, 2002; Eastman & Billings, 2001; Kuo, 2003).

Given that play-by-play commentary of fast sports burdens the processing capacities of commentators, processing imperatives tend to have linguistic reflexes, most notably grammatical strategies for postponing the subject of a clause (Green, 1980; Ferguson, 1983). The reason is that postponing the subject (horse or player) gives the commentator time to retrieve that name from memory. In slow sports and during color commentary conversations, when the game is not proceeding, such strategies are not as necessary and are not greatly in evidence.

Other features of sports commentary grammar have been noted such as copula deletion (Ferguson, 1983). These also seem to be a function of processing pressure.

Many sports commentaries tend to have associated suprasegmental and paralinguistic characteristics (Keller, 2003). The pitch of ice hockey commentary tends to rise and fall with the pitch of the spectator noise (Kuiper, 1996). Racing commentaries are droned or chanted (Kuiper, 1991). However, in breaks between overs in cricket in England, the play-by-play commentator and his color commentator partner sometimes sound as though they are having a conversation in a private club. Again such prosodic traditions are specific to particular sports and regions. Even the prosodies of a play-by-play commentator during the play and when the play is not proceeding are different.

Rugby Union Football

Introduction

We turn now to our chosen sport. Rugby football is said to have begun in 1823 when William Webb-Ellis, a pupil at Rugby School in England, ran with the ball during a football game. At the time, the rules allowed a player to pick up the ball but not to run with it. Running with the ball was an innovation. After the invention of rugby football, as with many sports originating in England, the rules of rugby union were codified and the game spread to the English colonies. In England it was and still is a game played and followed by the upper and upper-middle classes, association football (soccer) being the game played preferentially by the lower-middle and working classes. Rugby football is now played in many countries as a minor sport. In New Zealand, rugby football is a dominant sport and regarded as the national game.

Commentaries of the same game involving different commentary teams may now be heard on various media including commercial radio and television. It is now also to be heard in Maori on the Maori television channel in New Zealand (Kuiper, King, & Culshaw, 2013). Broadcasts from different media thus provide opportunities to compare the effect of the medium on linguistic variation in commentary speech.

The Lineout

Rugby has what are termed "set pieces" governed by rules of what may and may not happen. Among these are set pieces to restart the game after it has been stopped. They include the kick off, which starts or restarts the game at its commencement, after the halftime break, or after points have been scored; the drop out, which is also performed by a drop kick, though not at the halfway line but at the defending team's 22-meter line; the scrum, which is a complex maneuver in which the ball is returned to play, usually after an infringement of some sort; and the lineout. In this paper we shall focus on lineouts for reasons that will become evident in the next section.

The events constituting a lineout are as follows. The position where the lineout is taken is determined by the manner in which the ball leaves the field. If the ball goes into touch (i.e., out of play unintentionally), the location corresponds to the place where this happened. The ball may also cross the sidelines intentionally. That can be done by kicking the ball into touch, which is one of the ways for gaining territory. Territory is gained or not gained depending on whether the ball was kicked from outside the 22-meter line when it was last grounded and also on whether the player who kicked the ball out of play was in an attacking or defending situation. Territory can also be gained if a penalty kick is awarded and the ball goes into touch.

Once the location of the lineout is established, players walk or jog toward it. This can take some time because the location of the throw in may be many meters away from where the ball was kicked. The touch judge stands at that point with a flag raised, indicating which team has the throw-in to the lineout. The team designated to return the ball to play by throwing it in chooses the maximum number of players who should be in the line. The maximum lineout contains all of the forwards. The minimum is two from each side. If there are more than two, the team that does not decide the number in the lineout may have fewer but must not have more than the number decided.

Figure 1.

Lineout positions. The relative positions of players at a starting point of the lineout are represented in this diagram. The team that has the throw-in is represented by the seven filled dots. And the unfilled dots are the other team's line. The dashed line on the left is the 5-meter line behind which players line up. The dashed line to the right represents another ten meters, beyond which lined-up players must not go until the ball is thrown. The other dots represent other players, including the halfbacks, who are the only players not lined up that can enter the area near the lineout. The black solid line represents the sideline. The sketch is not drawn to scale.

When the nature of the throw is decided, this may be relayed in code or in words. For example, a captain may consult with others and then say to the other teammates and the thrower, "We want it off the top."

In the lineout the two single lines of forwards must stand a meter apart, between the 5-meter and the 15-meter lines, which is parallel to the touchline at a right angle to it (see Figure 1). Of the backs, only the halfback from each team may stand nearer than 10 meters from the lines. The others must be farther away.

The hooker who throws the ball in usually wears the number 2 jersey. His opponent, the other team's hooker, stands a little aside between the touchline and the 5-meter line.

After the players have lined up in the required configuration, the throw-in is taken.

The thrower stands out of the field of play and must not place a foot on it as he throws the ball between the two lines. The manner of the throw can be short to the front of the assembled players, long and high to those at the back, or high to a player in the middle of the line.

Only one player may jump up high to retrieve the ball. The player who jumps may be supported by other teammates, who may not lift him, only support him, in midair. The manner for doing this is prescribed. The player who takes the ball in the air may tap the ball away, bring it to ground level, or catch it with both hands and throw it to the halfback. If a front throw is played, the player at the front may elect to throw it back to the thrower. The thrower has some restrictions as to where he might be until the lineout is completed. This moment occurs when the ball leaves the lineout, either carried or thrown beyond the 15-meter line, or when a ruck or a maul develops where both feet of all the players leave the line of touch in the lineout,² in which case the rules of that play apply. The ball must be thrown straight; and if it is not, the lineout will be called again, upon which a penalty is applied.

From this description it is possible to determine the subepisodes of lineout play:

The sequence of subepisodes can be formalized in the following rule where the arrow represents "consists of," the plus represents "is followed by," the parentheses represent optional subepisodes, and the slash represents alternative subepisodes.

Lineout → Assemble + Throw + Jump + (Lift) + Take + Dispatch/Outcome

There are a number of lineout options:

• • Assemble: number of players (standard or short lineout)

• • Throw: long, high, short, front, back, and also the target player

• • Take: Two hands, tap, bring down, or move position for take

• • Dispatch: bring down, run, distribute

Aims of the Research

In this study, we will look at the effect of the medium on the message by comparing radio and television commentaries of the all the lineouts at five rugby international matches. It is clear from our outline in the introduction that the medium may have an effect on the speech of commentators, but exactly what that effect is in a particular sport and for a local commentary tradition is worth investigating. While Ferguson (1983) asserts that the differences between radio and television commentaries are slight, he also views them as therefore interesting. We will attempt to see how interesting they are and in what ways.

It might be asked why we focus on such a micro event and in only one country. The early study of commentary varieties tended to place them under stylistics and register studies. They also tended to be studied at a high level of generalization (Biber & Finegan, 1994; Crystal & Davy, 1969; Ferguson, 1983). Today there is a clearer realization of the locality of sports commentary. We have suggested that a useful framework within which to make that case is that of genre studies (Askehave & Swales, 2001; Bakhtin, 1986; Martin, 2001). We take it that lineout commentaries are a microgenre (Simons, 1978).

We now define two central terms. Play-by-play and color commentary can, for our purposes, be distinguished as follows. In play-by-play commentary the commentator is relaying the events that are taking place at the time. Such commentary is directly event driven. In color commentary the commentator is not doing this. Instead he may comment on any number of things. In association football, for example, he may mention that the player is to change clubs next year. In cricket he may comment on the weather forecast or on where seagulls are sitting in the stands. A second commentator often provides color commentary, but in association football generally one commentator provides both play-by-play and color commentary during the game while an expert, often an ex-player, will join the play-by-play commentator at halftime. The two will then both provide color commentary.

In terms of their linguistic features, play-by-play and color are distinguishable by a number of formal, or textual, features, notably by the density of formulae in the former, often also by their prosodies (Kuiper, 1996). Play-by-play commentary tends to be chanted or droned, particularly in the commentary of fast sports. In the following commentary excerpts, play-by-play sections are italicized.

As Michael Foley will have the throw to the lineout. Will it be one of the most crucial lineouts ever? It's taken by Cockbain quickly off the top.

Defensive lineout throw under pressure, New Zealand. It's taken down by Giffin—taken down by Giffin for Australia. That's the second one they've flogged.

Giffin from the back. Runs forward and takes uncontested ball. Beautiful stuff, Australia.

On the basis of the discussion above, we will test a set of hypotheses relating to possible differences between radio and television media utilizing the lineout. They are:

1. 1. Radio commentators will produce more commentary of lineouts than television commentators. This will be measured in the number of lineouts that pass without commentary, words per lineout, and clauses per lineout.

2. 2. Radio commentators will devote more linguistic resource (as measured in clauses) to less central aspects of the lineout than television commentators.

3. 3. Because viewers can see what is happening, television commentary will devote more commentary (as measured in clauses) to color comments.

4. 4. Because viewers can see what is happening, television commentators will produce more evaluation than radio commentators.

Method

The corpus of utterances for this study consists of transcriptions of all the lineouts broadcast for three games in the 2000 season of the trinations tournament, played between Australia, New Zealand, and South Africa, and for two further test matches in the following two years. There were 142 lineouts in the five games. The television corpus represents the commentaries of two play-by-play commentators, and the radio commentaries are those of two other play-by-play commentators. Transcriptions are parallel in that each lineout has a transcript from a radio and television commentary.

In order to gauge the effect of the medium, radio or television, on the way in which sports are broadcast, we have selected a number of textual parameters of increasing detail. The first parameter is the number of lineouts that receive commentary, since we know from previous research on ice hockey face-offs (Kuiper & Haggo, 1985) that in television commentary not all such episodes receive commentary. The second is the number of words and clauses per commentary used for each lineout. Third, we present comparative data from the way in which the subepisodes of the lineout are or are not given commentary. Fourth, we compare the way in which color and play-by-play are used in the two commentary genres. Fifth, we look at the kinds of affective modifications made by each commentator in each medium. Finally, we examine the use of formulas in the commentaries in the two media.

All lineouts were therefore coded for the following parameters:

• • words per lineout

• • clauses per lineout

• • clauses devoted to play-by-play per lineout

• • clauses devoted to color commentary per lineout

• • clauses devoted to every subepisode of each lineout

• • positively evaluative words or clauses per lineout

• • negatively evaluative words or clauses per lineout

Results

Our hypothesis is that since television watchers can see a lineout for themselves, whereas radio listeners cannot, there will be lineouts that are not provided with commentary on television. That is the case.

Uncommented Lineouts

Television: 16

Radio: 0

Our hypothesis is that since radio commentators must fill the time available with talk, they will produce more commentary than television commentators, who can indulge in short periods of silence since their listeners are also watchers of the video.³

For television commentary the mean length of utterance (MLU) per lineout is MLU 9.5 words, with a standard deviation of 6.1, whereas for radio commentary it is MLU 16.7 words, with a standard deviation of 7.7.

A second measure is the number of clauses per lineout commentary. We have taken an expansive definition of clauses to include small clauses in which copular verbs are absent (e.g., "Blackadder in the center").⁴

For the television commentaries there were 212 clauses for the 126 commentaries of lineouts (i.e., 1.7 clauses per commentary). For radio there were 395 clauses for 141 commentaries. (One commentary was indistinct and so has not been included.) There were therefore 2.8 clauses per commentary.

Our hypothesis is that since viewers can see what is happening for themselves, television commentators will provide less of the detailed description of what can be seen but more interpretative commentary, whereas the task of the radio commentator is to relate the events that the listener cannot see and therefore there will be less color commentary. That hypothesis is borne out. The proportion of color to play-by-play measured in clauses is about 1:2 in television commentary, whereas in radio commentary it is about 1:4.

There are a number of subepisodes in a lineout that might receive commentary, as we saw in the description under The Lineout in the second section. Some of these are more significant strategically than others. The team that wins the lineout has a strategic advantage from having possession of the ball. The take episode is therefore strategically central. Second most significant is the throw since being able to take the ball is dependent on the accuracy of the throw and the call from the hooker relating to how the lineout players are to take the ball. The following commentaries are annotated showing the subepisodes in square brackets.

To the lineout. [throw] Off the top beautifully. [take] Albert van den Berg, he sets it. [take] There it is for Zwanepoel. [dispatch]

Shortened lineout. [assemble] This one going further back. [throw] Dragged down by ah Flavell. [take]

Giffin up beautifully. [jump] Two-handed take. [take] Foley in the wrap around at halfback. [dispatch]

To gain a measure of the weight devoted to a particular episode, we counted the number of clauses devoted to that episode. The way to read the tables below is therefore to see that there are, in the case of television, 126 opportunities when the assembly of the players might have been provided with a clause or two of commentary, but there was no clause devoted to this episode. For the 141 opportunities for radio commentators to provide commentary of the assembly episode, however, 17 clauses were devoted to the assembly of the players at the lineout. The percentages are the percentage values for the same measure. Thus 12.06% represents the 17/141.

Table 1.

Clauses Devoted to Subepisodes of the Lineout

Table 2.

Ranking Episodes in Terms of Linguistic Resources Devoted Them

It is clear that in both radio and television commentary the ranking of what is conceived to be the relative significance of the subepisodes of the lineout is the same. However, strategically peripheral episodes such as the assembly and the jump are more likely to be given commentary on radio than on television. In neither medium is the work of those who lift the jumper granted even the occasional comment, however important it might be that the jumper is held aloft to take the ball.

As well as providing directive play-by-play, commentators also have the capacity to evaluate what they have seen. Such evaluations are either positive or negative. This might be done by a single adjective or adverb or by a clause. Since radio commentators take more time to describe what they are seeing, it may be that they have less time to evaluate what they see. There may also be differences in the degree to which evaluation is positive or negative. In the excerpts below, the evaluative constituents are italicized.

And Eales takes the New Zealand throw easily at the front. Great stuff from the champion lock forward.

Oliver to throw to a full lineout, and it's Maxwell who goes up; two hands secured nicely.

And it's knocked back on the New Zealand side, not straight. That's a good call, and that's a terrible throw from Oliver and Cribb.

Table 3.

Positive and Negative Evaluation by Television and Radio Commentators

The data suggest that both television and radio commentators are equally likely to make evaluative comments and are alike in their degree of preference for positive evaluation of what they are seeing.

As well as having discourse rules of the sort we have investigated above, where the order of what is said is, in part, determined by the external events that are being seen by the commentator and, in part, by the conventions of the commentary genre, football commentaries are also formulaic (Lord, 1960). That is, much of what is said is constructed of formulas, these being lexicalized phrases that are indexed for particular purposes. This can be illustrated by a selection of formulas from the lineout data. In Table 4 the following conventions are observed. A slot (i.e., an empty position that is filled by a noun phrase of a particular kind) is indicated in caps, so FORWARD is used for the name of a forward. Optional constituents are in parentheses and alternative constituents are separated by a slash.

Discussion

The data presented in the previous section suggests a number of significant findings for the study of sports commentary genres and their text types. First we have shown that although there may be a genre of sports commentary, or of rugby football commentary, commentary on television and radio are significantly different. The differences in the form of the text type can in part be ascribed to the context, as is often supposed with genres (Bauman & Briggs, 1990). In this case the context is not that the commentators are watching different games or from different positions. The game is the same game and the commentators are in a commentary box. The difference is that one is broadcasting to an audience who are listening on radio, the other to an audience who are watching the game on television. It is not that these two different audiences create different expectations on commentators. It is easy to imagine the wires being literally crossed and the television audience listening to the radio commentary. We know of some viewers who prefer the radio commentary and thus leave the audio on their television off and listen to the radio commentary of the game they are watching. However, we have shown that the medium does have some of the kinds of effects that one would predict.

Second, commentary is not just the descriptive relaying of everything that happens in a game. Commentators have a wide experience within a game and thus know which aspects of it are significant. In the case of the lineout, as seen by someone who is not a rugby aficionado, it may seem as though the spectacular lift of an already tall and usually large forward player would be the most notable event. However, that is not what commentators mention, because it is not as significant as who takes the ball, since that is strategically (although not balletically) important. The often desultory arrival of the forwards (many of them need a breather), at the point where the lineout is to take place, although it takes a good deal of time when the lineout is far down the field from where the play was previously, does not often rate a mention since it is not strategically significant.

Third, there comes a point when it is difficult to tell whether these preferences and the way they are expressed becomes part of a tradition (i.e., when the form of the text type becomes conventional) (Guenther & Knoblauch, 1995). In this study we have taken a synchronic approach, but it would take a diachronic survey over the time that rugby has been broadcast to ascertain how long the way rugby commentaries as we have described them have been the way to provide commentary on rugby lineouts. It is not that the two explanations are at odds. The strategic importance of some subepisodes over others is clear, but it may not always have been precisely so. It may also be that commentaries of rugby lineouts in other parts of the world are significantly different since genres constantly evolve (Pare & Smart, 1994).

Fourth, there are likely to be aspects of the rugby commentary genre that take precedence over the effect of the medium in which the commentary is broadcast. The ranking of the significance of the subepisodes of the lineout, the rate at which evaluative comment is made, and the ratio of positive to negative comment, all appear to be medium independent, and thus we can presume that these are features of rugby commentary oral traditions in New Zealand. The formulas of rugby lineout commentary are also likely to be part of a New Zealand tradition of rugby commentary. The reason for supposing this is that rugby commentary was initially wireless radio commentary and only later became television commentary, so the formulas of radio commentary are likely to have been imported into television commentary. Furthermore, many commentators

Table 4.

Lineout Formulas

who work in television broadcasting worked in radio broadcasting earlier. They would have brought their formulas with them.

Conclusion

This case study has shown that there are text-type features of rugby union commentary in New Zealand that are sensitive to the medium in which the commentary is broadcast, while other text-type features are features of New Zealand rugby football commentary traditions. Some of the medium-specific features appear predictable—for example, the amount of commentary, the amount of color commentary compared with play-by-play, and the likelihood that particular episodes are provided with commentary. However, other features, such as the amount of evaluation of the events a commentator provides, cannot be predicted on the basis of the medium of the commentary. This suggests that sports commentary traditions must each be examined, each in its own right. That is to be expected with oral traditions.

Further research into sports where both radio and television broadcasts are available may corroborate these findings. Another subject for further investigation is the effect of the speed of the sport on the differences between radio and television commentary language. The prediction would be that the faster the sport, the smaller the commentary differences, since with fast sports both television and radio commentators must keep up with the game.

In terms of its practical repercussions, if the findings of this research are sustained by studies of other sports, then it would seem that radio commentary furnishes a more useful apprenticeship for television commentary, since radio commentary fosters the provision of greater detail and greater fluency on the part of the commentator.