- Introduction
Due to Covid-19, those living in the United States have been required to wear a facial covering for most activities. This has become a continued norm in our society. We wondered how this influenced communication in society, specifically focusing on how those with accents that are non-California American accents are affected. How has this impacted those who are living in a city where their native language is not spoken? Are nonnative accented individuals inclined to use more gestures and eye-contact during speech to make sure they will be better understood?
Another interesting question brought up is how those without an accent change their speech and listening behaviors for those with an accent. How has the interaction between the two communicators been influenced? Would they be more likely to ask more clarifying questions? Do they move closer towards the person with an accent to better hear them? Or when they are speaking with somebody with a different accent, do they enunciate their speech more? Do they speak slower? We are hoping to answer these questions, or shed some light on the interaction through this study.
1.1. Background
The Covid-19 pandemic has dramatically altered day to day life and communication. Face masks used to prevent the spread of the virus have possibly hindered our previous level of interpersonal communication. Face masks take away the ability to observe facial gestures in communication, specifically the lower and middle areas of the face which play the biggest part in emotional recognition (Mheidly, 2020). Many people rely on the crucial social cues relayed through facial expressions in communication even without the barrier of a facial covering. For instance, much of communication is a combination of interpreting body gestures as well as interpreting the language. According to Segal et al. (2020), nonverbal communication can play 5 roles: repetition, contradiction, substitution, complementing, and accenting. We believe that hindering an aspect of nonverbal communication can affect these roles which can make communication more difficult. For instance, with the mask mandate, more people have been unknowingly pushed to use prolonged eye contact (Mheidly, 2020). Thus increases feelings of uncomfortableness and anxiety, but hinders emotional engagement between people as well as communication between the individuals (Mheidly, 2020). In narrowing down our research we focused on learning how native Mandarin speakers have been impacted by facial coverings taking away the visual facial gestures used in communication.
Communication is already quite difficult for those with an accent. Many feel deep frustration with communicating because they aren’t being understood or perceived differently (Heblich, 2015). All throughout the internet there is an abundance of tips and tricks to both be understood as well as to be able to understand accents. In addition to this, speech pathologists offer therapy to reduce accents although wearing a facial mask has just added more difficulty. Gluszek and Dovidio (2010) found extensive research that shows that within English speaking communities there is quite the stigma around those with non-native accents. Thus, this impacts the interaction between speakers, creating a dialogue where both speakers find it difficult to come to a conceptual understanding in language exchange. In conjunction with the shift in communication due to wearing masks because of the Covid-19 pandemic, we felt that studying masked social interactions with the additional factor of accents would be a unique test that could hopefully provide insight into our participants’ personal experiences.
Our paper will discuss the interaction between Mandarin speakers and Californian English speakers. We will be testing our participants in two environments one with a facial covering and without and one outdoors. The find the difference tasks will be the same with a facial covering as it is without but of different images, allowing us to see if there are differences in communication with a facial covering and without a facial covering. Our without facial covering task is done to receive a baseline on a typical interaction between the two linguistic groupings. We expect to see more clarifying speech and body language between the Californian and Mandarin paired participants while wearing a facial covering. We do not expect any significant differences in clarifying speech and body language between paired groups of similar linguistic backgrounds (i.e. The Californian pair and the Mandarin pair). This hypothesis stems from the inclination of people to more positively interact with those with similar accents, own-accent bias (Bestelmeyer et al., 2014). If they are communicating with somebody that has a similar linguistic background we believe they will more efficiently interact and establish common ground more quickly than with those that do not share similar linguistic backgrounds.
1.2. Questions & Hypotheses
1.2.1. H1: California x Mandarin: We believe that the Californian participants
will tailor their accent to be more easily understood by the Mandarin
speakers. We believe that the restriction of a facial covering will cause
California participants to slow their speech and enunciate their words.
Mandarin participants will also change their typical speech patterns. We
expect them to also slow their speech and more carefully enunciate. With
both we expect high levels of clarifying questions and/or constant repeated
speech outside of normal standards. These results will exceed the ‘typical’
behavior that was shown in the baseline task.
1.2.2. H2: Mandarin x Mandarin: We expect interactions between participants
who are both Mandarin speakers to be similar to our baseline observations.
1.2.3. H2: California x California: We expect interactions between participants
who both have Californian accents to be similar to our baseline
observations.
- Methods
2.1 Participants
There were a total of five participants; 2 native Californians and 3 speakers of Mandarin. All participants within the study were recruited from the University of California – Santa Barbara and due to time constraints, were selected based on convenience sampling. The average age of our participants is 19 years and 5 months. In addition to age, 60% of our sample identified as female while 20% identified as male, and the final 20% identified as non-binary. The predominant ethnicity amongst our sample was Asian with 60% of our participants selecting this answer. Other ethnicities from our sample included 40% being White or Caucasian, and 20% being Hispanic or Latino, with one participant identifying as both White/Caucasian and Hispanic/Latino. The majority of our participants were in their junior year of college with 60% making up our sample. In regards to our remaining participants, 20% were freshmen, 20% were sophomores, while none of our participants were seniors in college. Our non-native participants have not been living in the United States longer than 10 years, with the average amount of time spent in the United States amongst our Mandarin-speaking participants being 4.3 years. Specifically, the average amount of time spent in California amongst our three Mandarin speaking participants is 4 years and 8 months. From our Mandarin-speaking participants, our first participant said they spoke less than an hour of Mandarin a day, our second participant said they spoke only three hours of English a day and the rest of the time spoke Mandarin, while our third participant said they spoke seven hours of Mandarin a day. Our participants are currently undergraduate students at the University of California – Santa Barbara and in addition to this, participated in a questionnaire that gave further information on their experience with English, which allowed us to conclude that the non-American Mandarin-speaking participants within the experiment have excellent English competency.
2.2 Materials
2.2.1 Surveys
Participants needed access to the internet in order to complete a 5 minute online post-survey that asked questions based on their demographics, personal experiences in their respective cultures and native language, and residency within the United States. Beyond the initial consent question informing students of a brief overview of the experiment and stating the right to withdraw from the study at any time, in addition to consent of use of video, audio, and photography for the purpose of the study, participants were asked questions about basic demographics such as age, gender, ethnicity, and year in college. In addition to this, our questionnaire was meant to gather data relevant to our participants’ experience as non-natives or natives to the United States, asking questions such as “What country/countries do you have citizenship in?”, “How long have you lived in the United States and California?”, “Do you speak any languages outside of English?”, “What is your native language?”, and “How many hours per day do you speak Mandarin?” .
All responses were given the option to be recorded in short answer text. Our remaining post-survey questions were asked as follows: “Do you feel that wearing a facial covering makes it more difficult to be understood?”, “Do you feel that others are more difficult to understand with a facial covering?”, and finally, “Do you knowingly change your speech in any way to help facilitate communication?”. Once again, all these responses were given the option of short answer text. This questionnaire was essential in attaining a more detailed understanding of our participants as a whole, as well as to gain a more complete overview of the differences between the cultures they come from.
2.2.2 Diapix Task
Our participants will be engaging in a Diapix task, in which they must communicate with one another to find the differences between the images presented to them (See Figures 1 through 6) without seeing the other’s image. The Diapix Task was originally developed at Northwestern University in the study Van Engen et al. (2010) and used in the Wildcat Corpus, a project of the Speech Communication Research Group at Northwestern University (Van Engen et al., 2010). In a study by Baker et al. (2011) they were able to further design the former Diapix, known as the DiapixUK, which extends the task into 12 comprehensive picture pairs for participants to use. This procedure has been used with the ages 18 years old to 39 years old with both native and non-native speakers in various languages and is useful in showing interactions such as the frequency range of communication, speech rate, and specific word associations with the pictures within the Diapix in order for the participants to say certain vowels for speech analysis (Baker et al., 2011). Arguably one of the biggest strengths of the Diapix task is its ability to create a natural communication flow between two participants, with the possibility of adding role giving (i.e. instructions or taking charge in one participant) or to create an almost equal amount of conversation to be analyzed if desired (Baker et al., 2011). The Diapix task has also been observed to allow participants a free range of vocabulary to describe the objects within the pictures presented to them, which is helpful when working with a variety of backgrounds within participants and for multiple tasks.
2.3 Procedure
Originally we prepared a 10-15 minute long task for participants to complete, however due to a time constraint we had to shorten all our sessions to six minutes. In a future experiment beyond our pilot, we would extend the time in order to have longer and greater material to analyze. Our baseline pairing were our two Californian native participants, and preceded by this were both our Mandarin speaking participants, and then finally one Californian participant and one Mandarin participant. Every session between our respective pairs used English as the language of communication. In addition to this, we split up our pairings into two types of sessions; using the Diapix task without facial coverings and with facial coverings in respect to the ongoing pandemic. All pairings were recorded outside in six minute sessions per Diapix task. Each participant received a new partner to minimize the impact of established common ground.
The task in each scenario will be the same. They will be doing a find the difference task in which one participant will be asking each other questions about the photo to spot the difference between the two. They may not look at each other’s photos. The images chosen were from diapix: there are 3 versions of the image of a street we will give them and for the second part of the experiment we have 3 images of a beach. The ordering of if they perform the baseline task or the experimental task first will be random.
Outside Diapix task without facial covering (See Figures 1, 2, & 3):
- Californian participant x Californian participant (Baseline pairing)
- Mandarin participant x Mandarin participant
- Californian participant x Mandarin participant
Outside Diapix task with facial covering (See Figures 4, 5, & 6):
- Californian participant x Californian participant (Baseline pairing)
- Mandarin participant x Mandarin participant
- Californian participant x Mandarin participant
2.4. Data Collection
For each experimental task we took video of the participants
(with consent) to evaluate their performance. We tracked the number of clarifying
questions asked, how often they had to repeat themselves, hesitations, and observed their body language and speech patterns. In addition to this, we also recorded the amount of sentences spoken, and the duration of their speech to compare the amount of turn-taking spoken between both participants.
2.5. Analysis:
If there are high amounts of clarifying questions and/or repetition in the
experimental task compared to the baseline task between the Mandarin x California
participants then we can confirm our first hypothesis. Also, if speech patterns and pace
shift once participants have a facial covering then we can confirm our first hypothesis.
We will be using the videos of the participants as our tool to analyze these aspects.
Experimenters will watch the videos and manually count hesitations (‘um’ and ‘uh’),
clarifying questions, moments of loss of words, and self-correction. We will also count the amount of sentences each participant spoke and the duration of each turn they take. Some results were disregarded due to overhead planes where we could not hear speech clearly.
3. Results
3.1. Survey:
3.1.1. Part 1: Survey was to gain demographics and consent. Survey
questions are attached to the appendix.
3.1.2. Part 2: Survey was meant to collect the participants’ opinions regarding the
topics discussed in our experiment: the adaption of communication
while wearing a face mask. Questions are attached in the appendix.
3.2. Video observations:
3.2.1. Beach 1: This was the paired California accent participants. This
task was done without the facial covering. We recorded 2 instances
of correctional speech, 17 hesitations (predominantly ‘um’ and
‘uh’), and there were no instances of clarifying questions nor
repeated speech. Overall both participants had a low tone and
spoke at a normal pace throughout the video. There was
minimal eye-contact however, nodding gestures were used
frequently. We recorded 44 sentences from California 1 adding up
to 145 seconds. For California 2, we recorded 45 sentences adding
up to 161 seconds.
3.2.2. Beach 2: This was the pair of the Californian participant and the
Mandarin participant. This was also without facial coverings. We
recorded 5 instances of correctional speech. 26 instances of
hesitation and 2 moments of asking for clarification. There also
were 2 moments of ‘loss of words’ which was not seen in the first
beach task. We also observed moments of confusion, longer pauses
in between speech, and body language directed towards each other.
Another interesting observation was that the California participants
essentially led the task and spoke more than the mandarin
participant. It also seemed that they were speaking a bit louder than
the paired California participants. For California 1 we recorded 61
sentences, lasting 142 seconds. For Mandarin 1 we recorded 41
sentences, adding up to 101 seconds.
3.2.3. Beach 3: This was the last task without wearing a facial covering.
This last one was between two Mandarin speakers. We observed 4
moments of asking for clarification, 7 moments of hesitation, and 5
moments of self-correction. One of the participants had body
language directed towards the other but that was not matched.
They also spoke quietly compared to the other participants. We
recorded for our second Mandarin participant 55 sentences, adding
up to 68 seconds. Then recorded 53 for Mandarin 2, lasting 67
seconds.
3.2.4. Street 1: This was the Califorian pair’s task with the facial
covering. We recorded 3 moments of self correction, 1 moment of
a loss for words, and one clarifying question. Interestingly, there
were 34 moments of hesitation. Similar to the first pair
run-through, there was very limited eye-contact, turn-taking, and
body gestures that convey understatement such as nodding.
California 1 had 60 sentences, lasting 124 seconds. California 2
had 58 seconds lasting 113 seconds.
3.2.5. Street 2: This task was between 2 newly paired mandarin speakers
wearing a facial covering. We recorded only 3 moments of self
correction, 4 clarifying questions, and 5 moments of hesitation.
There were noticeably longer pauses between speeches. Overall,
there were not any body gestures being used, other than both
participants were leaning towards each other. We recorded
Mandarin 1 having 45 sentences, bringing a total of 79 seconds.
Mandarin 3 spoke 50 sentences, adding up to 93 seconds.
3.2.6. Street 3: Lastly, this task was the mandarin speaker and californian
pair, both wearing facial coverings. There were 3 clarifying
questions and 18 moments of hesitation. Again, predominantly just
the Califorian participant spoke. Both participants were leaning
towards one another however, they used relatively low voices.
Also, their body language conveys understanding through repeated
nodding. There was very minimal eye contact. California 2 spoke
72 sentences, totaling 92 seconds. Mandarin 2 has 50 sentences,
adding up to 70 seconds.
4. Discussion:
Our study aimed to find the effects and patterns of speech within non-native Californians and native Californians in relation to our present communication with facial coverings. After reviewing the data we collected, we have come to the conclusion that the results we received were inconclusive. Between the masked and unmasked trials we conducted, the observational results were not different enough to come to a meaningful conclusion. While the data does not support our first hypothesis of participants tailoring their language for clarifying communication when wearing a facial covering, it does support our second and third hypotheses of the pairings of Mandarin speaker x Mandarin speaker and Californian accent x Californian accent matching closest to our baseline observations. Likewise, according to the responses we received for the survey, the majority, 60%, of participants felt that they both shifted their speaking patterns with communication and found it difficult to understand others or be understood while wearing a mask.
If we were to continue on with this project the points we would take into consideration moving forward include addressing the issues we had while running the pilot. This includes having had to reduce trial times due to time constraints as well as the possibility of anxiousness due to being recorded. It can also be considered confounding the limited set of participants and languages we used. Therefore, to better execute our experiment, if we moved forward with it, we would consider expanding our sample size of participants, as well as possibly the type of language speakers we would use in a more randomized method of recruitment. We would also consider including warm-up sessions before each trial to better prepare our participants for the tasks and allow them to familiarize themselves with each other.
Acknowledgements
A huge thank you to the UCSB Engaging Humanities Program, especially Elina Salminen and our class mentor William Chavez for being so helpful throughout the duration of this course. In addition to this, our acknowledgement would not be complete without the overwhelming guidance from our professors Argyro Katsika and Kim Yasuda for providing us with resources critical to the development of our experiment and being overall wonderful facilitators and instructors during our time in INT-136PA. We would also like to thank our five participants for giving us their time to engage in this experiment.
This experiment was not possible without the collaborative efforts of Giselle Woods, Hailey Apolloni, and Sarah Young. All aspects of our project regarding the conception, experimental design, and development of our project (Questions, background/research, & hypotheses) were created between Giselle Woods, Hailey Apolloni, and Sarah Young. All credit to recruiting our participants is dedicated to Hailey Apolloni and Sarah Young. The experiment was conducted by Giselle Woods and Sarah Young, however it took place near Hailey Apolloni’s residency. The data distribution was spread between all three of us, with Sarah Young distributing the data, Hailey Apolloni analyzing the data, and Giselle Woods computing and organizing the final data.
Appendix:
Survey:
Part 1:
Initial consent question & a brief overview of our experiment was provided at the beginning of the survey.
Demographic Information
- What is your name?
- Age
- Gender (If you would rather not disclose, please put N/A)
- Ethnicity & Race
- Year in college
- What is your citizenship?
- How long have you lived in the United States?
- How long have you lived in California?
- What is your native language?
- Do you fluently speak any other languages outside of English? If yes, which?
- How many hours per day do you speak English? How many hours per day do you speak Mandarin?
Part 2:
- Do you feel that wearing a facial covering makes it more difficult to be understood?
- Do you feel that others are more difficult to understand with a facial covering?
- Do you knowingly change your speech in any way to help facilitate communication?
Link: https://docs.google.com/forms/d/1DD9WAbAK2J_hhnKdhGNMsWuJSdOWCfE_hOQBbzxY3dI/edit
Figure 1
Diapix Task of Street 1A & 1B Used Between Both Californian Participants With Masks
Figure 2
Diapix Task of Street 2A & 2B Used Between Mandarin Participant 1 & 2 With Masks
Figure 3
Diapix Task of Street 3A & 3B Used Between Californian Participant 1 & Mandarin Participant 3 With Masks
Figure 4
Diapix Task of Beach 1A & 1B Used Between Californian Participants Without Masks
Figure 5
Diapix Task of Beach 2A & 2B Used Between Mandarin Participant 1 & 3 Without Masks
Figure 6
Diapix Task of Beach 3A & 3B Used Between Californian Participant 2 & Mandarin Participant 2 With Masks
Figure 7
Visual Representation of the Amount of Speech Patterns Observed From Session 1 (With Mask) & 2 (Without Mask) With Both Californian Participants
Figure 8
Visual Representation of the Amount of Speech Patterns Observed From Session 1 (With Mask) & 2 (Without Mask) With Both Mandarin Participants
Figure 9
Visual Representation of the Amount of Speech Patterns Observed From Session 1 (With Mask) & 2 (Without Mask) With Both Californian and Mandarin Participants
Figure 10
Diapix Trial Recordings:
https://drive.google.com/drive/folders/1OvMuvRG-yt27tlFnlMyqjJ63a8LnZi4S?usp=sharing
Works Cited
Baker, R., & Hazan, V. (2011). DiapixUK: task materials for the elicitation of multiple spontaneous speech dialogs. Behavior research methods, 761–770 https://www.phon.ucl.ac.uk/project/kidLUCID/diapix.php
Bestelmeyer, P. E., Belin, P., & Ladd, D. R. (2015). A Neural Marker for Social Bias Toward In-group Accents. Cerebral cortex (New York, N.Y. : 1991), 25(10), 3953–3961. https://doi.org/10.1093/cercor/bhu282
Gluszek, & Dovidio, J. F. (2010). The Way They Speak: A Social Psychological Perspective on the Stigma of Nonnative Accents in Communication. Personality and Social Psychology Review, 14(2), 214–237. https://doi.org/10.1177/1088868309359288
Heblich, S., Lameli, A., & Riener, G. (2015). The effect of perceived regional accents on
individual economic behavior: A lab experiment on linguistic performance, cognitive
ratings and economic decisions. PLOS ONE, 10(2).
https://doi.org/10.1371/journal.pone.0113475
Mheidly, N., Fares, M. Y., Zalzale, H., & Fares, J. (2020). Effect of face masks on interpersonal communication during the COVID-19 pandemic. Frontiers in Public Health, 8. https://doi.org/10.3389/fpubh.2020.582191
Segal, Jeanne., Smith Melinda., Robinson, Lawrence., & Boose, Greg. (2020). Nonverbal Communication and Body Language. Retrieved from https://www.helpguide.org/articles/relationships-communication/nonverbal-communication.htm
Van Engen, K. J., Baese-Berk, M., Baker, R. E., Choi, A., Kim, M., & Bradlow, A. R. (2010). The wildcat corpus of native-and foreign-accented English: Communicative efficiency across conversational dyads with varying language alignment profiles. Language and Speech, 53(4), 510–540. https://doi.org/10.1177/0023830910372495
