Hacking Verbal Communication Systems

Ryan Alexander

Ryan Alexander

#developer-experience-dx

#ruby-on-rails

#algorithms

Hacking Verbal Communication Systems

by Ryan Alexander

In the talk "Hacking Verbal Communication Systems" at RailsConf 2019, Ryan Alexander explores the challenges of verbal communication within modern development teams. He emphasizes that conventional conversational flow often leads to ineffective communication, especially in group settings and remote interactions. To address these issues, Alexander introduces a simple system of hand signals inspired by the Occupy movement and adapted from Quaker traditions. These signals aim to improve conversation dynamics by enhancing flow control and consensus modeling.

Key Points:

Importance of Conversation: Conversations are exchange processes where ideas must interact, not merely be presented in a one-way manner.
Flow Control Mechanisms: The speaker discusses how conversational flow is subconsciously managed by all participants but can become confusing with cultural differences and personal communication styles.
Occupy Movement Hand Signals:
- Alexander explains eight hand symbols used for effective communication.
- Four symbols are for flow control (e.g., wanting to speak next, contributing to the current topic, seeking clarification, and signaling that someone is meandering).
- Four symbols pertain to consensus modeling (e.g., express agreement, disagreement, the need for a break).
Race Conditions: Interruptions and pauses create race conditions that can hinder communication. By utilizing symbols, participants can express their desire to speak without interrupting the ongoing dialogue. This helps to prevent misunderstandings and promotes inclusivity.
Inclusion of Marginalized Voices: Alexander introduces the 'progressive stack' concept, which prioritizes the voices of those who haven’t spoken, ensuring everyone contributes to the conversation.
Neurodiversity Considerations: The hand signals can help neurodiverse individuals engage more effectively in conversations, as they often find body language difficult to interpret in real-time.
Implementation Tips: Alexander suggests using facilitators to guide discussions while also adopting these hand signals. Ensuring everyone understands the system is crucial for its success and accessibility.

Conclusion:

Ryan Alexander concludes by reiterating the adaptability of conversational algorithms. By integrating these hand signals into conversations, teams can cultivate better communication habits that eventually render the formal system unnecessary, allowing for smoother interactions. He stresses the importance of making these systems inclusive, as everyone brings value to the conversation.

00:00:20.660 Okay, so hello everyone! This talk is about hacking verbal communication systems. If you're not here for that talk, you might be in the wrong place.

00:00:25.320 I want to do a quick bit of housekeeping. First of all, my name is Ryan Alexander and my pronouns are he and him. I am the lead backend developer for a cross-functional Ruby on Rails team based in London, but I grew up right here in Minneapolis, which is why I don’t sound much like I’m from London.

00:00:37.650 In fact, I went to South High School and then I attended the University of Minnesota, where I received a highly technical degree in music theory and composition.

00:00:59.520 How to get ahold of me: You don’t need to capitalize the RNA part; that’s just for clarity. You can find me on Twitter and email. I’ve put a hashtag in the corner that you can use if you want to discuss this talk on social media.

00:01:13.440 I believe both the slides and a video of this talk will be available about a month after the conference. They are going to handle that and, even better, I mean they are being live-streamed, so if you want to reference them, please do. I encourage all of you to give me feedback, as this is only the second time I’ve given this talk.

00:01:35.940 The first time was about three years ago at the Lead Developer Conference in London. I thought it would be a one-off thing, but the feedback I received was so encouraging that I decided to apply to give this talk again. I’ve included a lot more content this time and I’m diving deeper into several areas that I didn’t get a chance to cover before.

00:02:04.590 I would really appreciate your feedback, as it means a lot to me and is probably true for every speaker you see at the conference. Additionally, I want to express that I truly value being at a conference with a strong code of conduct; it’s important to me. If there’s anything in my talk that feels unwelcoming, I want to know about it so I can avoid making the same mistake again.

00:02:24.420 If you have feedback on how things could be better or anything like that, I would love to hear it because I believe this talk has helped people, and I want to continue to do so without anything getting in the way.

00:02:44.730 All right, I would like to talk about how to improve our conversations. According to Google, conversation is defined as an exchange. This is the key word that I want all of us to pay attention to: exchanged.

00:02:52.349 Conversations are not meant to be one-way exchanges; they involve multiple ideas. If we want to bring diverse sets of expertise from our teams to bear on our goals, individuals don't just need to be heard; their ideas must interact. There needs to be a back-and-forth which often occurs through conversations.

00:03:20.099 What I specifically want to talk about is how that back-and-forth occurs. We're going to dive into two related systems: the first is flow control, by which I mean who gets to talk, when they get to talk, and how that decision is made. The second is consensus modeling, which involves how we figure out what the group wants.

00:03:37.739 Starting with flow control, how does everyone involved in a conversation determine who will speak and when? The short answer is that we all do it collectively, but the person currently talking usually wields the most decision-making power. In fact, we are doing that right now.

00:03:59.280 This is a one-way conversation and I appreciate the irony of discussing a one-way conversation about having multiple ways to converse. However, feedback is happening, as I’m observing your physical cues. I’m noticing whether you are looking at me, smiling, or nodding along, which is helpful.

00:04:29.159 The longer answer involves our capacity to think of potential hacks—things we can improve. Everyone in the room is running an algorithm in their head about who should talk and when, and then communicating their opinions through body language, short signals, and facial expressions.

00:04:51.300 If that sounds like a mess, it often is. It generally works for most people, sometimes better for some than others. I think an example might help illustrate this, so let’s go back to Ireland.

00:05:12.360 When I was 19 years old, I saved enough money to travel around Europe for three months on a budget. A friend who had moved to Paris offered to host me, so the stars aligned. I traveled throughout the UK, France, the Netherlands, and Germany, but something odd happened in one quiet village in Ireland called Bunclody.

00:05:38.100 I have to clarify that this wasn’t the case for all of Ireland but specifically this one village. It was there that I first started to realise our flow control systems. This was a long time ago, so if you decide to visit Bunclody, it may have changed since then.

00:05:58.500 I found that I couldn't hold a conversation. I'd begin talking but would soon stop abruptly. This happened repeatedly during the night. Interestingly, the locals had no issue communicating with one another.

00:06:14.220 It was clearly a combination of me and them, and I couldn't initially figure out what was occurring. Eventually, I realized it was about conversational flow control. I would sit in the pub, attempt to chat about my travels, but whenever I tried to elaborate, I would inadvertently stop.

00:06:35.670 The other person engaged in a different flow control system that I was not accustomed to. While I was talking about my travels, they gave me signals to indicate I could continue, like a sharp intake of breath, but since I interpreted it differently, I would stop.

00:06:55.320 This miscommunication created a choppy dialogue, like someone trying to drive a stick shift after only using automatic. It became evident that we held different algorithms for these interactions.

00:07:12.340 In Bunclody, the differences were startlingly clear, but often, conflicting systems just keep functioning without us noticing. Sometimes, we think we have the same algorithms, but we really don’t.

00:07:29.380 When we do have differing systems, it can interrupt the flow of conversation. It's also to be noted that there may not even be agreement on what constitutes an optimal conversation. To be fair, even if we agreed on the same algorithms, each of us brings our own subjective importance to the conversation.

00:07:53.960 We generally think what we have to say is important! Now, quick confession: this is not an original hack. Most of us have been employing a version of this technique since childhood. However, for various people, it feels inequitable.

00:08:15.320 I’m going to convince you why it’s worth implementing. Let’s look at the hand signals adopted by the Occupy movement. I’ll provide some historical context about them. The Occupy movement did not invent these signals; they were actually adapted from the Quakers, who have used them in large group meetings for a long time.

00:08:40.160 The Occupy movement took up these symbols when faced with local ordinances that hindered their use of artificial amplification. They needed a way to organize protests effectively. Hence, they created a system with a human megaphone in which one speaker would talk, and people close enough would repeat for those further away.

00:09:00.300 However, this disrupted natural conversational flow, and they sought a new way to facilitate interaction peacefully. There are four symbols for flow control and four for consensus modeling.

00:09:15.250 The first signal is the familiar raised hand, indicating that the individual wishes to speak next. The second signal signifies a desire to contribute to the current topic being discussed.

00:09:41.750 The third signal, which is my favorite, is a request for clarification. It indicates that the person would like to keep participating but doesn't understand something. If someone is using an acronym, for example, this signal communicates a desire for clarification without stopping the conversation.

00:09:58.420 The fourth signal expresses a request for the speaker to stay on topic, indicating that they might be wandering a bit. The other four symbols represent feelings—expressing disagreement or agreement, signaling enjoyment or opposition.

00:10:10.320 In professional settings, you rarely observe signals that require stopping a conversation. Yet, including these can serve as a safety net for those who might feel unable to express that the conversation has gone into uncomfortable territory.

00:10:25.000 It's also crucial to remember that your algorithms differ from those of others. You cannot predict what might prompt someone to want to exit a conversation. This also allows individuals the opportunity to have a positive cutoff.

00:10:46.750 If everyone in the room is signaling agreement, then it’s a cue to stop talking, which can be an excellent tool for those who may be prone to over-explaining or struggle to gauge others’ responses.

00:11:12.320 To demonstrate how these symbols can improve conversations, let’s examine some major flaws in our collective algorithms, namely race conditions and interrupting jerks. I am pleased I made this slide as I had previously used a more crude term.

00:11:35.150 In conversations, we often insert small or longer pauses to signal either a desire for acknowledgment or that we’ve finished speaking. However, at times, we skip these pauses, which can cause anxiety among listeners.

00:11:53.720 When we do not allow pauses, people feel an urgency to interrupt. This can create nervousness or irritation about the failure to find a suitable opening.

00:12:12.610 These interruptions can amplify anxieties, especially in public speaking scenarios where the unspoken expectation to return the 'stick' in conversation is often broken. People feel a sense of unease when they're unable to contribute.

00:12:35.340 These pauses we introduce can feel like starting pistols, marked by race conditions. Once initiated, even small latency can disrupt the flow of conversation, especially if you're connecting with others remotely.

00:12:53.400 You notice this often when two groups of people communicate over remote connections, especially in office settings where visual cues and delays can hinder smooth exchanges.

00:13:11.880 The symbol hack mitigates this issue by removing that startup pistol. You can raise your hand the moment you realize you want to enter the conversation without interrupting those currently speaking.

00:13:29.210 This also allows you to listen more intently to the content being conveyed, rather than being preoccupied with anticipating a pause to jump in.

00:13:47.040 However, even if everyone hears that starting point simultaneously, some may be faster to react. Moreover, some people won’t wait at all, leading us to address the problem of jerks.

00:14:07.020 To clarify, not everyone skilled at interrupting is a jerk, but interrupting is generally inconsiderate. When someone jumps the gun, others often don’t bother trying to signal.

00:14:27.080 What we see is a current speaker’s interruption, negating the efforts of those waiting to speak, disproportionately affecting marginalized voices. Again, this is where hand signals can offer assistance.

00:14:49.390 Using hand signals means you don't need to interrupt someone else to enter the conversation. They also allow everyone to visibly acknowledge those waiting to speak, enhancing the effort towards consensus modeling.

00:15:09.470 When you see how many people are waiting to contribute, it also gives a sense of the group’s needs. This method requires you to classify the nature of your interrupt request. For instance, if someone needs clarification, they can be prioritized over someone wanting to shift topics.

00:15:38.790 Many individuals already instinctively prioritize contributions based on visual cues. However, skilled ideas do not inherently grant their creators the ability to fluidly enter conversation.

00:15:57.860 Interrupting is a learned skill, and it’s one that some are rewarded for throughout their lives while others are penalized, leading to biases along sexist and racist lines.

00:16:11.800 Allowing someone who is good at interrupting to dominate should not imply that the conversation will improve. In fact, it likely perpetuates inequities.

00:16:29.320 Let’s discuss consensus modeling. A challenging aspect here is that in any conversation, there are usually two layers of consensus at play. The first layer is obvious—the agreement on the topic itself.

00:16:43.680 Then there's a second layer, which isn’t as straightforward. This is where hand signals can enhance communication by clarifying whether it’s acceptable to continue discussing the topic or if it’s time to shift gears.

00:17:06.160 Typically, both sets of consensus signals rely on body language and tone. When discussions reach a pivotal point, explicit statements often emerge to guide communication.

00:17:22.720 These symbols function like subtitle tracks for our physical language cues and become crucial as cultural interpretations of gestures can vary widely. For instance, in Bunclody, these symbols greatly help when cultural meanings diverge.

00:17:44.020 Additionally, neurodiverse individuals may find it more challenging to decipher body language in real time, making it harder to incorporate feedback organically into discussions.

00:18:02.840 If you're working in technology, it's important to foster inclusivity for neurodiverse team members. If you assume you don’t encounter diverse perspectives, you may very well be contributing to this issue.

00:18:27.540 Now, in wrapping up, I'd like to cover some points regarding how to implement these techniques properly. You can certainly do this with an explicit facilitator, like a scrum master, who serves to shepherd conversations.

00:18:44.640 These symbols can support a facilitator's role, easing their tasks and enabling more individuals to share that responsibility. It's also essential to keep the principle of inclusivity in mind while ensuring that discussions are equitable.

00:19:10.640 Additionally, attendees should be mindful about employing the progressive stack to prioritize marginalized voices. Essentially, you focus on hearing from those who haven’t had the spotlight yet.

00:19:36.320 When participants raise their hands, it’s often a good indicator that they are willing to engage. However, you must remain aware that some individuals might not want to participate, and calling them out could have adverse effects.

00:19:58.840 Every system has limitations. Importantly, if someone in your group cannot see the raised hands, or if you’re on an audio-only connection, these symbols won’t serve their intended purpose.

00:20:24.460 Nevertheless, if you have a chatroom connected to your audio setup, entering versions of these signals within the text chat still makes sense as people can utilize them there.

00:20:46.700 There are also several bots developed to help manage this type of interaction. These can ensure that when someone raises their hand, they are brought to the forefront of the list if they haven’t spoken yet.

00:21:02.600 These hand signals represent one of many systems that enhance conversational flow. While other systems offer similar benefits, I appreciate the simplicity of hand signals, as they introduce inclusiveness without unnecessary complexity.

00:21:22.790 As a final note, I’ve been inadvertently hacking you all during our discussion. The subconscious algorithms we've explored are adaptive, much like us. By discussing how different types of interrupts should be prioritized, you begin to integrate these ideas into your understanding of conversation.

00:21:49.560 Questions arise—should you seek clarification, or is it time to shift to a different topic? It becomes easier to express these inquiries when familiarity is established.

00:22:20.190 Remember Bunclody? The next morning, I woke up and could converse without issue; my brain adapted and found the flow. Most individuals can adjust similarly.

00:22:43.730 As these adaptations occur, some of these techniques may become unnecessary. However, it's crucial to remember that newcomers may lack the context or understanding of your established conversational patterns.

00:23:02.240 Be ready to revert to explicit techniques to assist them in the process. Avoid letting your new system become an exclusive secret handshake leading to confusion and disconnection.

00:23:23.250 Your algorithms may differ significantly from new participants. Guide them effectively through these processes, providing signposts and direction, as everyone brings value.

00:23:42.880 Lastly, be intentional about your culture and you'll reap the benefits of enhanced collaboration. Thank you for your time, I appreciate it.

00:24:21.310 Thank you.

RailsConf 2019