This is an observation that goes back to at least Cicero.[1]<p>Cicero on the primary goal of oratory:<p><i>"As, therefore, the two principal qualities required in an Orator, are to be neat and clear in stating the nature of his subject, and warm and forcible in moving the passions; and as he who fires and inflames his audience, will always effect more than he who can barely inform and amuse them..."</i><p>Cicero describes the problem the OP reports:<p><i>"But let us return to Calvus whom we have just mentioned,—an Orator who had received more literary improvements than Curio, and had a more accurate and delicate manner of speaking, which he conducted with great taste and elegance; but, (by being too minute and nice a critic upon himself,) while he was labouring to correct and refine his language, he suffered all the force and spirit of it to evaporate. In short, it was so exquisitely polished, as to charm the eye of every skilful observer; but it was little noticed by the common people in a crowded Forum, which is the proper theatre of Eloquence."</i><p>Nuanced communication not working at scale, 2100 years ago.<p>[1] <a href="https://www.gutenberg.org/cache/epub/9776/pg9776-images.html" rel="nofollow">https://www.gutenberg.org/cache/epub/9776/pg9776-images.html</a>
Whenever articles like this about communication crop up, I like to remind people of Wiio's Laws[0] - from 1978:<p><pre><code> 1. Communication usually fails, except by accident.
2. If a message can be interpreted in several ways, it will be interpreted in a manner that maximizes the damage.
3. There is always someone who knows better than you what you meant with your message.
4. The more we communicate, the worse communication succeeds.
5. In mass communication, the important thing is not how things are but how they seem to be.
6. The importance of a news item is inversely proportional to the square of the distance.
7. The more important the situation is, the more probable you had forgotten an essential thing that you remembered a moment ago.
</code></pre>
Just to reiterate, these laws are over 40 years old, predating any kind of social media or internet-based communication - and they're still 100% valid.<p>[0] <a href="https://en.wikipedia.org/wiki/Wiio%27s_laws" rel="nofollow">https://en.wikipedia.org/wiki/Wiio%27s_laws</a>
As another example, Amazon teams communicate product launch requirements via a future press releases including a FAQ (per description in the book "Working Backwards"). Its a communication intended for the masses with a built-in disambiguation addendum.<p>Our natural languages uses incremental inquiry to disambiguate context as opposed to using strong protocol. In "Working Backwards", it's the communicator's job to solicit questions from co-workers via pain-staking detailed reviews in meetings ("Bezos scrutinizes every single sentence"). I think of it like constructing a representative survey of ambiguity, and then putting answers in the FAQ that help increase clarity. The more detailed and representative your survey, the more helpful your questions/answers will be to communicate nuance.<p>With regard to disambiguating through protocol, Organizations evolve jargon to increment protocol, which probably increases semantic alignment somewhat as group size scales. If you read about the history of language, the Rebus principle created protocols of formal alphabets; protocols like grammar gave us formal writing rules. Protocols like TCPIP let our computers talk. Protocol creates more rigid commitments for communication, but also increases potential semantic alignment. As a thought experiment, if we learned to dynamically and deliberately develop jargons en masse, it might create the channels to disambiguate context and communicate nuance at scale.
Dan Luu is right that public messaging -- and company-wide internal messaging -- tends to be bone simple and incomplete.<p>What's interesting is that the nuances don't completely fade out of site. They exist in quiet and sometimes quite intricate underground conversations. I've joined organizations where it was howlingly clear that the official messaging was not the way the company really ran.<p>That invites the question of whether it's worth staying long enough (and being bold enough) to get drawn into the nuanced underground dialogue, too. Sometimes yes. Sometimes that's quite exciting and makes the job more interesting and more durable.<p>Other times, it's just too hard to wiggle into that circle. Or that circle has its own evasions and power struggles. In those cases, it's easier to meet the basic formal requirements of the job, enjoy the extra time to have a rich life outside of work -- and think hard about what kind of next job would be better.
We live in an attention economy, both outside and inside companies. The rules that apply to B2C marketing largely apply inside companies as well.<p>Despite that we still have people that assume “I sent an email and I’m important therefore everyone got the message”. Try running those emails through some tool like Mailchimp and you’ll probably find less than 40% even opened the email, let alone read beyond the first paragraph.<p>I’ve done a lot of organising events for engineers inside companies where there are like 500+ engineers. You need email, slack, calendar invites and more to get people paying attention. And often they’re paying more attention to LinkedIn than what’s happening on the “inside” … you can run campaigns on LinkedIn that target your own people…
One of the comments for the thread.<p>> I see this for every post, e.g., when I talked about how latency hadn't improved, one of the most common responses I got was about how I don't understand the good reasons for complexity.<p>> I literally said there are good reasons for complexity in the post<p>It feels like this kind of half baked point scoring reply is just a risk of posting on the internet. I'm sure I've been guilty of it at times too, and I think forums like HN or Reddit encourage it.<p>So you end up having to be overclear in a way that hurts your message.<p>For example, in just my previous post in another thread, I was talking about how I felt IMAP and SMTP support was important for a mail provider. However, I felt that if I just left it at that, some pedant would come yell at me about how IMAP and SMTP are not secure protocols since they're plaintext. So I wrote out IMAPS and SMTPS to ward off that kind of pedantry.<p>But I'm still at risk of someone else wanting to score points indicating that actually, IMAPS and SMTPS isn't a thing. And they'd be sort of right, IMAPS and SMTPS are colloqial terms for their corresponding protocols over TLS, but you won't find an IMAPS spec, and if you look in the IMAP RFC, IMAPS is not something that is mentioned.<p>I don't know how you fix this.
Nuance is hard to convey in groups, but I believe that *a small part of the problem is a lack of design*. Many peoples' eyes glaze over when they see a wall of text in an email and they just skim rather than read. Some simple things to enhance communications can be the following.<p>* Use a few bullet points to put attention on the main points you want to convey.<p>* Without going overboard, use a tasteful amount of graphic design (bolding one key sentence or whatever).<p>* Break up a giant nuanced email into sections.<p>* If something is critical, make it visual: a picture, explainer video, or an infographic can be really useful for something key.<p>This is harder than it looks. A quote attributed to Mark Twain is "I didn't have time to write a short letter, so I wrote a long one instead." It's a lot easier to go overboard than to distill what needs to be conveyed into the core elements.
This is why there must be separate private and public conversations. A comment that is appropriate for exploring nuances in a small group is inappropriate for wider communication to a large audience. Social media's blurring between small group conversations and publicly broadcasted conversations removes this distinction.
In the IT industry this scenario is fantastically common:<p><i>Here is a proposal. It has this HUGE upside and this SMALL downside</i><p>response: <i>because of this HUGE downside, and the TINY upside, I reject this proposal</i><p>If they said "what if the upside wasn't as big as you stated but the downside is larger than you stated" at least you could discuss the evidence. But people love to leap on problems and devils-advocate them into the ground.<p>You see this all the time in IETF mailing lists. I'm not talking about nit-picking during working group last call on a standard, thats justified. People who simply want to be contrarian, take a devils-advocate stance, leap on any stated downside and on the premise its the proof, destroy the original idea, irrespective of the relative merits pro and con.<p>So, "we should move to Postgres because of its support of IPv6 and JSON" dies on "but the sheer amount of code we have in MySQL makes this untenable" -which is not a good argument, given the budget and willingness to incur the cost. It doesn't address the upsides of the move at all. Or "but we don't know all the places which use the old SQL forms" which is true, but presupposes we couldn't handle case-by-case the legacy calls into the old SQL binding, or find some way to uncover them.<p>The negative case arguments used, typically are shorthand for "I don't want to think about this"
I think the problem is that people, even smart people, lack critical thinking skills or don't apply them. It's not the scale that makes the nuanced communication not work, it's the scale that makes you notice it. You could have the same communications with smaller groups and you'd have the same results, you just wouldn't be as likely to get negative feedback indicating it.<p>It's also ironic to me that someone would try nuanced communication on Twitter, a platform whose very design discourages it. You can't do nuanced communication in 280 characters, but you can do vitriol just fine. So they do the tweet storm which turns off anyone who isn't incredibly interested in what you're saying.
This kind of doublespeak/mixed messaging with nuanced private conversations but dead-simple (or worse, spun) public statements is a double-edged sword.<p>Sure it makes sure that the x% of employees who don't get the nuance won't ask questions about it, but it prevents ambitious employees from learning how decisions are made at higher levels.<p>Understanding how the organization makes decisions can help you make decisions in your day-to-day work. Not to mention the fact that if you one day have aspirations of leading an organization, you need to understand how decisions are made. When over-generalized public statements are made, it not only conceals this information but corrupts it and can lead people to false understandings of how things are done.<p>Some of the best insight I've gotten about why my organization and my company makes the decisions it does have come from reading discussions from senior leaders in google doc comments. I wish I could be a fly on the wall for live meetings or private conversations.<p>In the optimal company, employees who don't want to have to grok the nuance would be able to trust the decisions of the leads. However all decisions should be made in the open so that those who do want to go to the effort of understanding something can learn.
I think 'communicate' is the wrong word when you address large group of people. You can't meaningfully communicate in that case.<p>You just produce content and the people are just consuming the content in whatevre way they please. Large percentage of people won't consume it in the way you wish.<p>Because of the disconnect what you want is way less important fir the result than what they want.
Napoleon was said to favor a tactic wherein he would bring in a lowly lieutenant to hear his orders, and repeat them back in their own words.<p>If the lieutenant could figure it out, then Napoleon could relay orders to his generals (who would in turn send orders to their subordinates and so on) with confidence that the meaning would not be lost on the battlefield.
Flat orgs are very popular right now, but isn’t it a huge benefit of a hierarchical organization with subparts that rather than the President of Azure getting on a VTC and telling the whole division that the goal is velocity he can explain to his reports (a small group) that they need velocity with reliability and they can explain to their reports (more small groups) and so on and so forth?<p>Yes, nuanced comms don’t scale so why isn’t the answer—-don’t require scaled comms?
Looking back, I work(ed) in a lot of environments where leadership tried to be nuanced. Or at least tried to communicate two or three equally important things.<p>Thinking about the outcomes of these with this explanation in mind does explain a lot.<p>Just not sure if this is a case of Confirmation Bias or a genuinely helpful way of looking at corporate communications.<p>Probably need more examples/data to better understand if he is on to something.<p>Nonetheless do I think it is a good framework as clearly communicating one thing and dropping the nuances would probably increase the likelihood that the content is being parsed as intended.
You can remove "at scale" and it's still true. You can possibly even remove "nuanced" as well. Try writing an email to just one person. If you put two questions in it you will only get an answer to one of them (if you're lucky).<p>Human language is an incredibly lossy form of communication. Practically the entire field of philosophy is a consequence of this.
I read it and felt kind-of bitter.<p>On one hand, yeah, may be nuanced message doesn't work at scale. However, when saying literally two things (vs one) became nuanced. I wound understand if it was a speech talking about a dozen of different things and their interplay... These were literally two things - create a solid product and let's move forward fast. That's it.<p>Also, why the hell whole hierarchy of middle management exist in such case? The only reason for it to exist is exactly ability to execute at scale (when things which are coming from the top are propagated properly).
Patrick Lencioni said this in The Four Obsessions of an Extraordinary Executive.<p>Two of his principles are:
1. Create Clarity.
2. Overcommunicate Clarity.<p>Executives always think they are overcommunicating (they hear themselves speak many times on the same subject so it gets super repetitive for them), but teams rarely have clarity.
While I won't disagree with the argument, I think the conclusion is flawed. If one must focus on velocity, and reliability is in opposition to velocity, then <i>how much</i> should one focus on one versus the other? It is not well defined, but that is ok. Since we are speaking to humans, not robots, reliability is not therefore completely disregarded--it becomes implicit, and deprioritized, but it is obviously still present to some lesser degree.<p>A good counterexample to the article would be Amazon's success with its leadership principles--much has been written about this, and I feel no need to repeat it here--or JFK's speech urging America to the moon, in which he spent significant time discussing the tradeoffs and sacrifices required to pursue the lunar landing, and in the end did not unilaterally decide to pursue the mission so much as he proposed a conversation about it and asked Americans to discuss the nuances and decide together. Nuance is possible at scale; it is a sad sign of the times that some now believe it is no longer possible.<p>If you haven't listened to JFK's speech, I strongly urge you to take a listen, and compare his measured, collegial tone to the tone of our politicians today.<p><a href="https://www.jfklibrary.org/asset-viewer/archives/JFKWHA/1961/JFKWHA-032/JFKWHA-032" rel="nofollow">https://www.jfklibrary.org/asset-viewer/archives/JFKWHA/1961...</a>
> A number of companies I know of have put velocity & reliability/safety/etc. into their values and it's failed every time.<p>I have a 'slow is fast' mantra and it's definitely something that a lot of people misunderstand, willfully or otherwise.<p>I've often shrugged it off as the fact that going fast is exhilarating, while the effort of 'making the change easy' starts to sound dangerously like discipline. Perhaps I've downplayed the fact that A->B can sound an awful lot like A & B.
Counter-examples: Harry Potter, Marvel, etc. High bandwidth, low latency (often days or less), unlimited scale (most of the planet), high retention.<p>Nuance will be pruned at scale if inessential. It’s a feature, not a bug.<p>The popular stories are [perceived as essential] for survival and propagation.
Semi-related:<p>The amount of complaining about "dumb management" amongst software engineers is truly out of control I find.<p>Newsflash, usually your principal engineer and VP are a <i>lot</i> cleverer than you, on average. They're at the top of a process to weed out idiots and are paid a multiple of your salary.<p>I usually start from the position of "my VP is a smart and clever person" and work backwards, ALWAYS. It seems like this attitude is exceptionally rare though.
>Meanwhile, the only message VPs communicated was the need for high velocity. When I asked why there was no communication about the thing considered the highest risk to the business, the answer was if they sent out a mixed message that included reliability, nothing would get done.<p>The managers obviously didn't want to use nuance. They openly state that it would then enable those who they manage to negotiate with them, which is a power dynamic they'd rather avoid.<p>If management was sensitive to the feedback, they could intelligently guide the organization through the trade offs involved, instead of blindly charging forward into subsequent, totally avoidable, brick walls.<p>Software isn't as simple as making gears. Gears themselves are nowhere near as easy to get right as you think, by the way. When you produce gears, there are well known trade-offs in terms of manufacturing costs, maintenance requirements, and service life. None of that is true with software.<p>Because the best source of information is with the programmers, and the users of the software, any feedback paths between them should be encouraged, and listened to by management, not crushed to meet the quarterly goals.
IIRC Jonn Kerry's campaign for President was plagued by too much nuance in both his and the campaign's communications.<p>And the Romney critique as the "flipping Mormon" can also be seen as a rejection of nuance.
I keep most of my points extremely simple and avoid complex language unless I'm with a group of people whom I trust to understand subtlety. If you look at the history of my comments, you can see people often completely misunderstand what I say, often attributing the opposite opinion to the one I hold (or stated; I often don't state my personal opinion).
It's not just corporate communication.<p>Post on any internet forum about some minor drawbacks of Technology X, and the comments will instantly split between those who feel personally offended and must defend the sanctity of X, and those who take the post as proof that X must not be used at all. People love to think in black & white. People don't like to parse "A if (B and (C or not-D))". It takes much training and discipline to overcome the instinct for simplification.<p>You can clarify yourself and correct any misunderstandings if you're in a small group that shares a lot of context, but this quickly becomes impossible as the group gets larger. Even competent statesmen struggle to convey "A if B else C" in their speeches.
Of course, the real skill is in delivering a simplified message. "Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away." -Antoine de Saint-Exupéry<p>When you appreciate the nuance, how do you decide what to strip away?
On a related note, <a href="https://news.ycombinator.com/item?id=11370189" rel="nofollow">https://news.ycombinator.com/item?id=11370189</a> has an interesting paper worth reading.
Twitter threads are really awful to read through and, in this case, funny since it is a thread about communication.<p>Nothing really useful in the thread but the conclusion caught my attention: "Azure has, of course, also lapped Google on enterprise features & sales and is a solid #2 in cloud despite starting with infrastructure that was a decade behind Google's, technically."<p>This didn't happen because of any communication strategy from VPs to developers. It happened because MS is an enterprise company with a strong brand.
the times that I've observed nuanced comms break down at scale, it is usually because genuine conversation is not possible, for a variety of reasons ranging from the number of people who would need to be involved in said conversations to political games of power. so then it becomes the job of a million layers of middle management playing telephone to translate<p>but nuanced communication is also a skill that can be worked on, and certain classes of misunderstandings can be mitigated. dan calls this out in his post, but ironically a major one was missed in it as well: the resulting conversation seems to be entirely about individuals capability (based on IQ and other BS) to understand messages conveyed to them, rather than about the complex organizational dynamics that might result in someone being pushed into interpreting a message as something other than what it is<p>EDIT: I'd also note that nuanced comms are equally difficult in large orgs irregardless of the size of the group being spoken to. For example, I've had VP+ (a smaller group given that the comms are going upward rather than down the org chart) misinterpret technical findings presented to them. There's so much extra cognitive overhead inherent in interpreting messages when you're in a Machiavellian experiment (aka the modern bigco environment)
Look at how many details this guy (Carl Sagan) conveyed in his 15' speech: <a href="https://www.youtube.com/watch?v=Wp-WiNXH6hI" rel="nofollow">https://www.youtube.com/watch?v=Wp-WiNXH6hI</a> in a way that potentially everyone understood. I wonder if every explanation was like his, if nuance wouldn't be well communicated, even to large groups.
This is a really great thread and really resonated with my personal journey, although at a smaller scale. I’ve transitioned within a large enterprise from leading a deep & broad engineering team with about 50 people to a vertically integrated ops/technology team with about 800 people.<p>The “work” is challenging, but the communications is much harder than I expected. It’s difficult to actually say anything at all because nobody will perceive what you say the same way, and then the telephone game with amplify whatever insecurities or worries that folks have.<p>The hard part about flipping from “move fast and break things” to more order is knowing when the right time is to transition. The other hard part is that the official “communications” functions live in a different vertical, and getting them involved often makes things worse. So we get stuck getting engineers and interns to communicate with people.
It seems "group communication can't be nuanced" doesn't capture the situation he's describing at Azure. My hunch would be management saw reliability levels as a given, something that daily orders wouldn't change whereas velocity was something that daily order could change.<p>But just as much, this is a Microsoft division. That company has historically won by having more features sooner than competitors with the problem of the whole thing being a mess being a secondary consideration. Sure, in this case, the risk is they'll push so hard the whole thing blows up - but that doesn't mean you don't push as hard as possible because it's at least perceived that if you don't that, whether you blow-up or not won't matter (see the concept of technical debt, etc).
And lots of people will only ever hear what you said second hand. After all nuance has been intentionally stripped out and the message has been spun to serve the aims of the person interpreting it. “Journalists” do this all the time, but I’ve also seen this happen in the workplace.<p>When I was in grade school I remember my teachers used to constantly stress the importance of primary sources. And I didn’t really understand the importance until much later in life. Why do all that work when someone else has neatly summarized it for me on wikipedia? Now I basically have a rule where I don’t allow myself to have a strong opinion on something unless I’ve gone and read up on primary sources.
I don't think the nuance is lost. It's just that the audience has their own agenda. They might be fully aware of the "AND" and or "BUTS", but will focus on what they want that message to really be.
There must be multiple levels to this. At Apple, at least for a time, there was the overriding idea that "It should Just Work." That was supposed to translate to every part of Apple products. So clicking on stuff in the Finder should just work and AirPort WiFi should just work and so on. What exactly working meant was totally context dependent, but the additional detail that minimal effort and oversight should be required was broadly understood and teams that made products that failed that test were repremanded and reorganized if their products were not outright cancelled.
Reminds me of politics. Try having a nuanced public policy discussion. What's especially bad about that is it makes people identify more strongly with their respective sides and double down on more extremest ideas.
This is something that I'm learning to appreciate. My recent HN traffic spike (Woe of WebSocket) had many people missing the point.<p>I'm preparing to launch my SaaS, and I'm wondering do I start with the fun and cheeky marketing "Hey, this was designed for board games, but you can use it for so much more" OR do I pair with "Infrastructure designed for JamStack" (which is a bit of a lie).<p>I've started writing the Amazon style press release and FAQ to help me, but I'm excited to start the train on selling a crazy new platform.
This is very distressing to me. When I try to communicate using examples or analogies, people often get stuck on a particular example and try to solve that example.<p>Communication is very hard.
>when I joined Azure, I asked people what the biggest risk to Azure was and the dominant answer was that if we had more global outages, major customers would lose trust in us and we'd lose them forever, permanently crippling the business<p>That's a pretty good question to ask, at any job interview. I will use this tip! (won't ask it on the first interview, but it's pretty usefull for a second or third interview)
regarding the apparently competing values of reliability and velocity, try propagating these goals through different mediums. code quality can absolutely be ingrained into the culture. things like TDD, when/what to integration test, and encouraging PR reviewers to enforce these requirements. you can literally just run an in-house course/training for everyone and the aggressively follow through.<p>once you’ve got reliability deep-seated in the culture, then you can talk about velocity all you want without worrying that your teams will make confused tradeoffs. in effect, you’ve communicated the nuanced concept that “we move fast, but always within the constraints that provide for a reliable product”. most people aren’t consciously thinking about it that way (not a bad thing), but their behavior matches what you were originally wanting to convey with a nuanced message.
I have seen this extensively covered in "Made to Stick" by Chip and Dan Heath. I got this recommendation from HN itself. It was a great read.<p>If your organization is large, you create one actionable, concrete, clear message, and you stick to it, and you communicate it downwards.<p>Seeing it in action is Microsoft is not shocking.
Most people are dumb, and can't keep two separate ideas in their head. This feels correct, especially in aggregate.<p>Many of the failures of public policy in recent years stems from a failure to confront how stupid the public at large is.
This has always been kind of obvious to me. If you utter a sentence that requires any amount of interpretation to understand, then more people = more possible interpretations.
Nuanced takes more effort to process. Nuanced often comes ofc as having a smell (i.e., not a good one). When most everyone else is nuanced then simple and direct sticks out.
> Nuanced communication usually doesn't work at scale<p>"Nuanced communication usually doesn't work." seems more accurate. Being precise and clear is hard.
This is why memes are so powerful .. you can communicate a 2- or 3- part thought with a well designed meme so that eg 50 % will get it rather than just 5%
Good... article?<p>It's ironic that he's communicating this on Twitter, a medium where nuance is particularly hard to convey, and where the audience is especially prone to missing it. I'm sure that's not lost on him.<p>One thing to note: the part about it being less of a problem when people misunderstand an article on HN than if they misunderstand a business communication made me think of (one time) when the CEO of our company defined a new strategy based on an article on Product Lead Growth he'd read. Or rather evidently misread, since he neglected the most important parts. My conclusion is that these things are interrelated, and mistakes can compound.<p>I really do think that reading comprehension is one of those things everybody (especially STEM people) assume they're good at, but usually they're actually just terrible at it, and supremely confident about that. The same goes with clear writing, which (to me) is even harder.
Does that explain the level of nuance around public health policy?<p>Covid policy for example. Some countries take into account natural immunity (more nuance). Others don't and simply require everyone to be vaccinated regardless (less nuance). Some recommend or even require kids to get the Covid vaccine (less nuance - everyone take it), other countries recommend against (more nuance - some should, some shouldn't). They all have access to the same data. Is this simply reflective of a different approach in communication? Or a different level of confidence or respect for the population to grok nuance?
I disagree. I think people are smarter than the 'elite' give them credit for.<p>It's just easier to control people if you justify treating them like idiots.
We need to talk more about this.<p>That said, Organizations with professionals should be able to do nuance, at least a bit of it.<p>But the general public at large ... you're dealing with 'lowest common denominator' which is 'issues with literacy' and harder to grasp - very limited, care, attention span, and may not even be listening to the message - and may be getting misinformation from elsewhere.<p>Communicating clearly is a skill.<p>A lot of marketing people I believe have missed the message on this, every day I come across a new product and can't really understand what it does, the value proposition, who it's for, etc.. while at the same time there's tons of arbitrary marketing verbiage. Words matter.
If Dan is right (and I think he is), have a stiff drink nearby and then think about how his point applies to something complicated that is life and death, such as conveying information to the public about a rampant virus.
Multiple people are pointing out that posting this thread on Twitter is probably a mistake and can't figure out why he would do it: Dan (currently) works at Twitter.