0:00:15 | or a graph to everybody |
---|
0:00:18 | implement |
---|
0:00:19 | animal but student at cmu with justine |
---|
0:00:22 | and i'm going to describe our work on |
---|
0:00:24 | automatic recognition of |
---|
0:00:26 | social conversational strategies |
---|
0:00:29 | which contribute to building maintaining a sometimes destroying of lighting relationship us to specifically these |
---|
0:00:34 | conversation strategies are reported things like self disclosure shared experiences a prisons go on |
---|
0:00:43 | let's begin with the motivation of the talk |
---|
0:00:45 | a speaker of course you multiple conversation goes in a dialogue and contributions low conversation |
---|
0:00:51 | can often be divided into a |
---|
0:00:53 | the like those that one for robust functions those that will fill interaction functions like |
---|
0:00:58 | turn taking and those that fulfill a in that wasn't functions |
---|
0:01:02 | which manage the relationship between the interlocutor's over time |
---|
0:01:07 | in the category of all that fulfil these in that wasn't functions are a conversation |
---|
0:01:12 | strategies |
---|
0:01:13 | which a particular we do nothing |
---|
0:01:15 | and i don't have an impact on |
---|
0:01:18 | the relationship between the two individuals |
---|
0:01:21 | so in this well we propose a technique to model and automatically recognize these conversational |
---|
0:01:25 | strategies |
---|
0:01:27 | from like using multimodal information the we use |
---|
0:01:30 | well the visual and the vocal modalities of the speaker as well as the interlocutor |
---|
0:01:34 | in the current and the previous done |
---|
0:01:37 | and we believe that it's important |
---|
0:01:39 | i as more natural conversations with dialogue systems become part of you closed at like |
---|
0:01:44 | to believe that the martin for on advancing the capability of the dialogue systems not |
---|
0:01:49 | only do |
---|
0:01:50 | they convey information energy was moved interaction |
---|
0:01:53 | but also manage long-term interactions by building intimacy and rapport |
---|
0:01:57 | not just for the sake of companionship |
---|
0:02:00 | but at the more intrinsic part of improving task performance |
---|
0:02:06 | clearly then ugly propositional content and the interactive content does not suffice |
---|
0:02:12 | when a parent well we're what what's a computational model of so should all in |
---|
0:02:16 | task context |
---|
0:02:18 | and basically we have investigated one of the most important roles and it's one filled |
---|
0:02:22 | by so fast and that is to build the bond between two people |
---|
0:02:26 | a one that is strong enough |
---|
0:02:28 | to allow people to build trust are with another person are not case without within |
---|
0:02:32 | the to compute the agent |
---|
0:02:34 | also we thought of as one as rubber or |
---|
0:02:37 | all the feeling of connection and how many with another |
---|
0:02:40 | and the sentence human this work is to develop a dialogue system which can facilitate |
---|
0:02:44 | that in the wasn't balloons with users all interactions in a long time |
---|
0:02:51 | rubber have been shown to have a good effect in fields such as education and |
---|
0:02:55 | that was you should and in fact upright a local actually develop i-th the adding |
---|
0:03:00 | computation model would suggest how interlocutors manager or two using specific conversational strategies which for |
---|
0:03:07 | which one concern a intermediate goals of rapport |
---|
0:03:13 | the foundation well by spencer only actually conceptualise is the interpersonal interface at the desired |
---|
0:03:20 | to be approved of once a positive traits and reducing studies like brace |
---|
0:03:27 | what to have been based management |
---|
0:03:32 | a private possible that's that interlocutor the what time and to increase coordination and by |
---|
0:03:38 | adhering to be here expectations |
---|
0:03:40 | which are guided by a more source about don's in the beginning state of the |
---|
0:03:44 | interaction and when a did i get snow each other it's more at the mind |
---|
0:03:49 | body interpersonal norms of the interaction |
---|
0:03:53 | so i fast age gender but was norms maybe work was divided well on the |
---|
0:03:57 | data the other person be here expectations |
---|
0:04:00 | on the other hand |
---|
0:04:02 | shared experience |
---|
0:04:03 | also allows |
---|
0:04:05 | to increase correlation between the two people |
---|
0:04:08 | by because people getting next their common history when they are definitely shared experience |
---|
0:04:14 | but like cementing the sense that people are part of the same unifying two |
---|
0:04:20 | and finally to better learn about the other person usual attentiveness is an important role |
---|
0:04:25 | obviously in our own corpus that mutual attentiveness is of fulfil i the strategy of |
---|
0:04:31 | self disclosure |
---|
0:04:33 | i the relation to perceive these that will become more intimate in nature |
---|
0:04:40 | the goal of this work that you're the coolest one understand the very nature of |
---|
0:04:44 | these conversational strategies by correlating them with a multimodal cues and are a man better |
---|
0:04:50 | article question is to leverage that i was standing to automatically recognize these strategies |
---|
0:04:55 | it can be implemented in a dialogue system |
---|
0:05:02 | so our corpus |
---|
0:05:04 | is the reciprocal peer tutoring corpus which was collected from twelve american english speaking kids |
---|
0:05:10 | who interact there were five weeks in a total of sixty sessions |
---|
0:05:14 | on an algebra topic |
---|
0:05:17 | and are |
---|
0:05:18 | part was demonstrates that there's tremendous amount of rapport building in this your dream context |
---|
0:05:24 | and this is a context to study the attic so social interaction |
---|
0:05:29 | which also had a one week das talcum so the trying to solve the problem |
---|
0:05:32 | of algebra or five weeks |
---|
0:05:38 | let's move the method |
---|
0:05:40 | as a prior work on detecting similar dialogue phenomena such as that of a slower |
---|
0:05:44 | and so it's one violation has i dialogue act modalities in isolation |
---|
0:05:49 | or has focused on like slowly data driven approaches the for instance one way to |
---|
0:05:54 | quantify a violation of a social norm is to see wendy language is different from |
---|
0:06:00 | the rest of the language in the dialogue so for example use of a cross |
---|
0:06:04 | entropy value |
---|
0:06:05 | twenty five is |
---|
0:06:06 | in a local recapture |
---|
0:06:08 | a richer variety of the sub categories of these conversational strategies |
---|
0:06:13 | and the maybe that is |
---|
0:06:15 | we construct rely the annotated corpus we cannot rely extensively official views |
---|
0:06:21 | on like using psychology give some psychology one a stand what strategies contribute to interperson |
---|
0:06:29 | closeness and then we asked three to five human raters to annotate buddies |
---|
0:06:33 | and computing the reliability so self disclosure here |
---|
0:06:37 | in our work |
---|
0:06:38 | was defined as well but expressions are which are used by people to really aspects |
---|
0:06:42 | of them that's to the other so we can present it into two types which |
---|
0:06:46 | is a enduring states |
---|
0:06:48 | which will be long |
---|
0:06:50 | and intimate aspect the ones that which of course and user to a very important |
---|
0:06:54 | within the context of a conversation also that would be done in a couple of |
---|
0:06:58 | mike that's |
---|
0:06:59 | but it also be once upon proceed actions which are socially unacceptable actions |
---|
0:07:05 | which are you know way |
---|
0:07:07 | we have other people feel better than the colours in but like i didn't value |
---|
0:07:10 | of the pretest a result with those negative numbers |
---|
0:07:13 | rf in the shared experience |
---|
0:07:16 | is an important way of showing that are the two people in a dyad have |
---|
0:07:19 | known each other |
---|
0:07:20 | and the getting that some commonality |
---|
0:07:22 | so we got we differentiated into sharing experiences outside the experiment an inside experiment |
---|
0:07:29 | for praise we had board label pretty that are labeled praise so this is an |
---|
0:07:33 | example of a label rate which is |
---|
0:07:35 | a great job with those negative numbers |
---|
0:07:37 | but it also be something like good job affect |
---|
0:07:40 | and finally |
---|
0:07:42 | also nominations are basically behaviors which go against |
---|
0:07:46 | generally accepted |
---|
0:07:48 | or steering wheel behaviors |
---|
0:07:50 | and the first pass decoder but actions one source-normalization and in the second pass |
---|
0:07:56 | we differentiated these categories which was a breaking the three rules which could be doing |
---|
0:08:01 | it off task a talk during tutoring attribute based writing acts like criticising in solving |
---|
0:08:07 | a teasing |
---|
0:08:08 | i don't also be referring to one |
---|
0:08:11 | your own or others social modulation |
---|
0:08:14 | right now focusing on the need to work and so on |
---|
0:08:17 | discourse relations actually signals of the guided coming closer and no longer feel obligated |
---|
0:08:23 | to adhere to the norms of the larger bow |
---|
0:08:28 | this is an example of |
---|
0:08:33 | impact of self disclosure in one of the dyads where like even says what we |
---|
0:08:38 | want to be when you want to dwell which is the eliciting self disclosure and |
---|
0:08:43 | we use a they don't know yet than anyone us that i want to be |
---|
0:08:46 | a chef |
---|
0:08:48 | and then the data was on and |
---|
0:08:51 | you say that a lot of like seven is larger than a book you wouldn't |
---|
0:08:53 | be in the middle and then be lost like actually me and never know thinking |
---|
0:08:57 | of making the you channel with completely off |
---|
0:09:00 | from this idea being a chef |
---|
0:09:04 | but however |
---|
0:09:05 | e he goes onto aspect you know |
---|
0:09:08 | you channels will make money |
---|
0:09:10 | and then a few done say to use as you know if anything you are |
---|
0:09:14 | making one |
---|
0:09:15 | i will reminder of which would be fine |
---|
0:09:17 | so that all back and forth and mean elicitation of l one and |
---|
0:09:21 | a cellular which is done by the other person |
---|
0:09:25 | your some other examples of violation of social norms so the top one is i |
---|
0:09:29 | that a friend dyads which was in which was or seem to be in high |
---|
0:09:34 | rubber which is |
---|
0:09:37 | so you want exactly that you're and beat with that ut in the top interaction |
---|
0:09:42 | and once as you can do that that's the whole point |
---|
0:09:45 | you say that hey you are probably never do that and then once said that's |
---|
0:09:48 | why are you doing you it might so that you're smiles and we just as |
---|
0:09:52 | you almost |
---|
0:09:53 | my gosh we never the that ever |
---|
0:09:57 | so basically this what the friend i and smiling in one very important background that |
---|
0:10:02 | we found across it does not even when friends do was a limitation always preceded |
---|
0:10:07 | or |
---|
0:10:09 | with this might or might always smaller than one for some additional colours |
---|
0:10:12 | which is one of kind of hedging ugly these violations |
---|
0:10:16 | i and the bottom example is actually strangers what perceived to be in europe or |
---|
0:10:23 | so here we use as a next problem is exactly the same is my any |
---|
0:10:28 | then that's was that you get what the problem and then they don't you have |
---|
0:10:33 | and then p two with that you know who overlap and says that serious exactly |
---|
0:10:38 | so this that was perceived to be in europe or and that strangers |
---|
0:10:42 | was being a selection of social number not be the best idea |
---|
0:10:45 | when this to forming a relationship |
---|
0:10:51 | we didn't go to for |
---|
0:10:53 | we will behaviors which are independent variables in this study so we have it is |
---|
0:10:58 | we have smiles and head nods |
---|
0:11:00 | and where it is we have created a partner |
---|
0:11:03 | that what you were doing using it what we bought very doing |
---|
0:11:06 | and then using as their in the room |
---|
0:11:12 | so the next up here is to |
---|
0:11:14 | understand |
---|
0:11:15 | like what you |
---|
0:11:18 | if the user when the |
---|
0:11:20 | you these conversational strategies to that extent we first undersampled in on annotated a set |
---|
0:11:25 | of on these were conversational strategies to create about in the dataset |
---|
0:11:29 | and the non annotated utterances were randomly generated |
---|
0:11:32 | so the final corpus consists of |
---|
0:11:35 | a house an example the sentence larger i don't want to examine the fate experience |
---|
0:11:40 | one sixty seven example the phrase |
---|
0:11:42 | and around ten thousand five an example of violation of those wrong |
---|
0:11:45 | then what that the bra sixty |
---|
0:11:47 | interaction sessions which is |
---|
0:11:49 | sixty one and how far interaction sessions |
---|
0:11:54 | in the next step we explored observable in verbal and vocal behaviors of interest |
---|
0:12:00 | we are drawn from a quantitative analysis |
---|
0:12:03 | so we used to work on twenty five i'll be able to use of interest |
---|
0:12:07 | and then use all can smile twenty five some simple low-level descriptors |
---|
0:12:11 | related to pitch loudness and the vocal quality and assess whether the mean value of |
---|
0:12:16 | these features are significantly different |
---|
0:12:18 | in utterances |
---|
0:12:19 | that were annotated the modifications are the end of a not out of eight |
---|
0:12:23 | with a conversational strategy and the side effects a stochastic generalisability |
---|
0:12:29 | and finally for visual behaviors and nonverbal behaviors we explore whether there are all operating |
---|
0:12:34 | with these conversational strategies and they look at the altar accuracy |
---|
0:12:39 | d quadrants like people |
---|
0:12:43 | the based on the statistical analysis we select which might be more to use to |
---|
0:12:47 | include in a machine learning model so we have three sets of features are the |
---|
0:12:51 | first because that is basically were able which will and will |
---|
0:12:55 | use of the input in the current down |
---|
0:12:59 | and in addition to that we also added to capture some context we also added |
---|
0:13:02 | some type of words we select bigrams |
---|
0:13:04 | you part-of-speech bigrams and the word part-of-speech pairs |
---|
0:13:07 | a feature set to is the listener behavior basically |
---|
0:13:11 | so what is the visual behaviour of the listener when important using a conversational strategies |
---|
0:13:15 | that's we just the two |
---|
0:13:17 | and features that we use to capture more context around the users of conversational strategy |
---|
0:13:21 | so features entry is one thing to the goodwin the previous turn |
---|
0:13:24 | the what was is what will clean visible |
---|
0:13:27 | expression |
---|
0:13:33 | we used and to regularize logistic regression as the training of all the pure and |
---|
0:13:38 | the estimated informants using accuracy and accuracy over chance |
---|
0:13:42 | and then the competitors some standard a very basic machine learning algorithms |
---|
0:13:49 | okay so let's move on to the results |
---|
0:13:52 | the ones that you article goal |
---|
0:13:53 | all of on understanding the nature of conversational strategies |
---|
0:13:59 | here are the results for these statistical analysis of multimodal cues was the disposal first |
---|
0:14:04 | also we found that when students ref for so we found that students effort significantly |
---|
0:14:07 | more onto their possible constant during the disclosure and we gotta talk about what the |
---|
0:14:12 | likes and dislikes |
---|
0:14:15 | the new categories of positive emotion what the negative emotional it also had a i |
---|
0:14:19 | effect size |
---|
0:14:22 | also we standardized look very but of what ethnicity |
---|
0:14:25 | which form light of the intuition that when people reveal themselves you know not handy |
---|
0:14:29 | are honest way they are more |
---|
0:14:31 | i come below one what are able |
---|
0:14:33 | we did that this way to report any city and it had a higher rate |
---|
0:14:36 | as well |
---|
0:14:37 | for acoustic features we found |
---|
0:14:40 | a moderate effect size for loudness |
---|
0:14:42 | in this mode utterances |
---|
0:14:45 | and this |
---|
0:14:45 | so |
---|
0:14:46 | are examination of the corpus we often found that |
---|
0:14:51 | like speakers often not excited are when the disclosed in the dialogue like or twenty |
---|
0:14:55 | it is not something fun suppressing about themselves |
---|
0:14:58 | the of in spoken lower voice |
---|
0:15:00 | when they were talking only negative about themselves |
---|
0:15:02 | so it the variation in pitch was not significantly over the only the loudness |
---|
0:15:09 | for which will you be found that the four types of gave since my where |
---|
0:15:12 | significantly more likely to operate in utterances of the let's go to compare two nonsense |
---|
0:15:17 | words or utterances |
---|
0:15:19 | with using it partner |
---|
0:15:21 | which had the highest effect size |
---|
0:15:24 | from a similar analysis for the listener but a good look at those details in |
---|
0:15:28 | the paper |
---|
0:15:30 | of a shared experience we look that affiliation driving time orientation what's one the book |
---|
0:15:36 | which it was only used by the close to a index commonality with an been |
---|
0:15:41 | within a given time frame where all that we do we make some kind of |
---|
0:15:44 | affiliation with the conversation partner |
---|
0:15:46 | and it wasn't was to affect size for both of them |
---|
0:15:49 | and |
---|
0:15:50 | like first wasn't obviously |
---|
0:15:52 | had a high effective include rapid whatever of and cultivation about that |
---|
0:15:59 | so the north a visual cues was similar to that of the twist motion |
---|
0:16:06 | next we look that creates |
---|
0:16:07 | all systems brain one for |
---|
0:16:10 | well billboards vision that increases the interlocutor's hundred and perhaps that if a k c |
---|
0:16:15 | i will have a positive tone of voice is a very intuitive and the war |
---|
0:16:19 | the a positive effect is what that |
---|
0:16:24 | we also look at some of the acoustic features here and we had a negative |
---|
0:16:27 | effect i swear loudness actually so people ls lower when they raise the partner |
---|
0:16:34 | and of them or at side effect is what will be quality features |
---|
0:16:41 | finally for source-normalization we looked at different categories of all asked all other things belonging |
---|
0:16:47 | to social categories or |
---|
0:16:50 | there was no concerns |
---|
0:16:52 | and the was present of a class about their |
---|
0:16:54 | we also |
---|
0:16:56 | it can capture the intuition that some signals in the language |
---|
0:17:00 | a puzzle slow modulation |
---|
0:17:02 | would stem from just putting one student in that you roll but address the problem |
---|
0:17:06 | you in context where one of the cuban one of the beauty |
---|
0:17:08 | and the change does |
---|
0:17:10 | so |
---|
0:17:11 | we also do better look at the power drive there was a small it was |
---|
0:17:15 | significant what the effect i was small |
---|
0:17:19 | and finally listened via well which has found a in we use you that are |
---|
0:17:23 | forced wasn't you're |
---|
0:17:25 | to be an indicator of high stages |
---|
0:17:27 | and in with user diverse wasn't singular |
---|
0:17:30 | to be a good predictor of lower stages |
---|
0:17:32 | we bought and the twins ones do a lot so implementation then we just there |
---|
0:17:35 | are more likely to make three statements which involve others |
---|
0:17:38 | so for first wasn't rule |
---|
0:17:41 | that it was significantly higher in source-normalization utterances what effect it was a small |
---|
0:17:48 | for acoustic features |
---|
0:17:50 | we had a positive effect size for the which the fun |
---|
0:17:53 | the loudness and the vocal quality features |
---|
0:18:03 | for the visual cues that would say that wanting one additional thing there was significant |
---|
0:18:07 | in for some additional head-nodding are we not finding the previous conversations are used to |
---|
0:18:11 | speakers where |
---|
0:18:12 | a more likely to had not when they were doing a violation of a social |
---|
0:18:16 | norm |
---|
0:18:19 | so then using these features or inform |
---|
0:18:21 | from these days as if they actually |
---|
0:18:24 | what them in the machine learning model any found |
---|
0:18:27 | logistic regression to outperform the other basic machine learning algorithms |
---|
0:18:30 | and b |
---|
0:18:31 | the accuracy or johns |
---|
0:18:33 | ranged from sixty to eighty percent for detection of these for a categories |
---|
0:18:39 | maybe weekly likely just go to the most predictive features which are more interesting than |
---|
0:18:44 | t |
---|
0:18:44 | like accuracy numbers |
---|
0:18:46 | so |
---|
0:18:48 | so in features that one |
---|
0:18:50 | which is |
---|
0:18:51 | this people behave in the current one |
---|
0:18:53 | we found and because that is close they are then to their partner |
---|
0:18:57 | by gazing at them |
---|
0:18:58 | and head-nodding pre-emphasized what they're saying |
---|
0:19:01 | a did not get of their own on the part that's worksheet |
---|
0:19:05 | and first person singular responsibly predictive however the effect that the machine learning wanna picks |
---|
0:19:10 | up for first one thing with much less there |
---|
0:19:12 | competitive model would be to indicating the importance of normal this in |
---|
0:19:17 | well while doing his conversational strategies |
---|
0:19:20 | listeners on the other hand respond |
---|
0:19:24 | during the current done by head-nodding to communicate their attention and giving and the speaker |
---|
0:19:29 | but not of the worksheet |
---|
0:19:32 | and in the previous turn |
---|
0:19:33 | it but there is that's like it is my and not and how well |
---|
0:19:38 | no or loudness in voice |
---|
0:19:42 | a four shared experience some of the most pretty if features |
---|
0:19:46 | included using a their own worksheet like the speaker the less likely to get at |
---|
0:19:51 | their own watch it all the integrated worksheet that i could have lower schumann voice |
---|
0:19:58 | a however affiliation driving time in addition what would have only two categories got here |
---|
0:20:02 | that was really pretty of shared experience |
---|
0:20:07 | i listeners |
---|
0:20:08 | on the other hand exhibited be a bit like smiling or have to indicate appreciation |
---|
0:20:13 | of the content of the tall or anything you could one |
---|
0:20:17 | also them but that is not a more likely to be elsewhere or at the |
---|
0:20:20 | speaker why the speaker is doing a shared experience |
---|
0:20:23 | but we are less likely do not |
---|
0:20:24 | and b that their own worksheet |
---|
0:20:28 | and finally in the previous done on the part of the last like to smile |
---|
0:20:32 | and gaze at their own worksheet |
---|
0:20:34 | and have a lower loudness in voice |
---|
0:20:37 | if the partner to the next most one which had experience |
---|
0:20:41 | operators |
---|
0:20:43 | the most predictive features because doing a print was giving a the buttons worksheet |
---|
0:20:47 | which route indicative of directing attention to what the speaker what part is doing well |
---|
0:20:51 | breathing him |
---|
0:20:53 | head-nodding with a positive tone of voice |
---|
0:20:55 | perhaps emphasize the praise |
---|
0:20:57 | smiling perhaps as an indication of a general appreciation |
---|
0:21:03 | artemis again the potential embarrassment of race |
---|
0:21:11 | also |
---|
0:21:12 | we got features for the listeners |
---|
0:21:15 | included |
---|
0:21:17 | head-nodding or back channeling an acknowledgement |
---|
0:21:21 | and in the previous turn |
---|
0:21:23 | you partner |
---|
0:21:24 | was more like use my |
---|
0:21:29 | and finally for source-normalization we found that the most predictive feature |
---|
0:21:34 | from the speaker's behaviour in the parent and you're accusing any part or smiling with |
---|
0:21:38 | my head nodding |
---|
0:21:40 | and private actually found that |
---|
0:21:43 | smiling is not only hitting any gettable it's all the time easement |
---|
0:21:47 | a display of appeasement |
---|
0:21:49 | and it's signal that you're of attitude at between source normalisation |
---|
0:21:53 | which is more likely to |
---|
0:21:54 | probable forgiveness one the interlocutor |
---|
0:21:57 | so |
---|
0:21:59 | thing in the interest of time i just about one or two implications of a |
---|
0:22:02 | lock |
---|
0:22:05 | then as well |
---|
0:22:07 | we identified some regularity the social interaction |
---|
0:22:12 | and we use might be more because reflectees conversational strategies |
---|
0:22:17 | a and e that applicable across a wide range of the mean because this mapping |
---|
0:22:20 | you know more generally also can apply to your to bring as well as |
---|
0:22:24 | what things like about of the for clinical decisions of words one |
---|
0:22:30 | and that some of you might have seen yesterday of these findings have been integrated |
---|
0:22:33 | into always of the system call sara |
---|
0:22:37 | which takes input in real time |
---|
0:22:39 | to detect conversational strategies |
---|
0:22:42 | feed it into the rapport estimated to estimate acquired level of rapport |
---|
0:22:48 | reasons |
---|
0:22:48 | about the source light intent |
---|
0:22:50 | and then generates behavioral all the form of a lot along with the interactions |
---|
0:23:00 | this time limitations of the work was that we use the valence in it |
---|
0:23:04 | and we would like to work with a more natural distribution only on the contrary |
---|
0:23:08 | and deal with this but you could machine learning method which you don't methods |
---|
0:23:13 | and the other one piece of that when we look at multimodal features instead of |
---|
0:23:16 | looking at them in isolation better to exploit the dependency of the correlation between different |
---|
0:23:21 | of each of the temporal contingency so it can look at it people for that |
---|
0:23:26 | i don't triangles of like these findings to build rapport align you |
---|
0:23:33 | the finally in conclusion |
---|
0:23:35 | we learn the discriminating power in general activity appears features |
---|
0:23:42 | speakers |
---|
0:23:42 | just are not you results in a shot |
---|
0:23:45 | speaker is usually accompanied it is crucial |
---|
0:23:47 | information we had not anything other partner |
---|
0:23:51 | listeners do not but the about their gaze |
---|
0:23:56 | also shared experience because the less likely just by and more likely to of or |
---|
0:24:00 | the gaze |
---|
0:24:01 | meanwhile listeners my signal in coordination |
---|
0:24:09 | the so that they were and happens to justine al and a and b |
---|
0:24:14 | and i think that |
---|
0:24:15 | and also the what it really could be that would you put this work |
---|
0:24:27 | we have done for one question |
---|
0:24:30 | basically |
---|
0:24:36 | i have in that a question about the term conversational strategy so i know we |
---|
0:24:40 | use it i've used it my own papers to that i was listening to you |
---|
0:24:44 | speaker that and thinking gosh it sure implies some kind of conscious intentionality about how |
---|
0:24:49 | i'm gonna approach the dialogue and it's unlikely that that's really was happening |
---|
0:24:54 | so i wonder if you when you're colours there's had discussions about what the caller |
---|
0:24:58 | and what really alternatives to conversational strategy that you might are considered |
---|
0:25:03 | well i think one of the was things was |
---|
0:25:06 | like thinking in terms of like |
---|
0:25:08 | this the first part of speech acts |
---|
0:25:10 | and |
---|
0:25:11 | so speech acts |
---|
0:25:12 | so the different we wanted to start again speech that these in my understanding is |
---|
0:25:15 | that |
---|
0:25:16 | on a six sre scan span more than one speech or |
---|
0:25:19 | and it's |
---|
0:25:21 | it's more about the illocutionary force of the utterances that morlet pragmatic rather than the |
---|
0:25:24 | actual |
---|
0:25:25 | x amount take what the linguistic content |
---|
0:25:28 | so that one reason for not |
---|
0:25:29 | right quality that make the move or speech-act what whatever conversational strategy |
---|
0:25:33 | but what also actually i've seen some work including where you have a taxonomy of |
---|
0:25:38 | dialog category |
---|
0:25:40 | and conversational strategy is |
---|
0:25:43 | is perhaps |
---|
0:25:44 | using the more complicated within actually we are doing so it it's |
---|
0:25:47 | it's more it it's more like |
---|
0:25:49 | something which can be inferred |
---|
0:25:51 | rather than |
---|
0:25:53 | like and ready narrative clause level as we are doing what do not is like |
---|