While reading John Gruber’s review of the 2020 iPhone SE, this part caught my eye:

What this source told me is that while developing the iPhone X, members of the team would typically carry two phones with them: a prototype iPhone X they could use, but (of course) not while in the presence of anyone who wasn’t disclosed on the project, and an older iPhone they could use in front of anyone. These team members would spend time, every day, using both phones. They knew they were onto a winning idea with the new interaction design for the iPhone X when they started instinctively using the X‑style gestures on the older iPhone, and never vice versa. When a new design is clearly better than an old one, it’s a one-way street mentally.

I believed that then, but I believe it more now after spending the last week with the iPhone SE. I’ve used it exclusively for hours at a stretch and I never stopped expecting it to act like a post-iPhone‑X device. I swipe up from the bottom to go home or multitask. I expect it to wake up just by tapping anywhere on the display. I pull down from the top right corner expecting to see Control Center. I can’t stop doing any of these things unless I’m consciously thinking about the fact that I’m using an old-style iPhone. Even if I locked my personal iPhone 11 Pro in a drawer and touched no phone other than the new SE for a week or two, I still wouldn’t shake my iPhone X interaction habits unless I abandoned my iPad Pro too.

Once you get used to the post-iPhone‑X interaction model, there’s no going back. A week with the new SE has not shaken my belief that the X‑style interaction design is superior. Not one iota. 

I politely object.

When you get accustomed to a new way of doing something, going back to the old one is always an exercise in friction. Sometimes the new user interaction model is indeed an improvement over the old one. But once the new user interaction model is ingrained in your muscle memory, wanting to apply it even when you’re interacting with an older device, or with a device that still uses the older model, is only natural. But we shouldn’t confuse this naturalness with ‘better design’ — it’s a form of bias.

Even if the members of the iPhone team were using a prototype iPhone X together with a regular iPhone with a Home button, I’m willing to bet the prototype ultimately got more of their attention, since it was the one they were working on. So it’s understandable that “they started instinctively using the X‑style gestures on the older iPhone”. If you think about it, it’s more natural to want to apply a new swiping gesture to the screen of an older iPhone than to want to push a physical Home button on an X‑style iPhone that doesn’t have one.

When I was considering upgrading to a newer X‑style iPhone^[1], I did a similar thing: I spent some time carrying both my iPhone 5 and an iPhone X that was kindly lent to me for research purposes. Interacting with both phones was generally a confusing experience, gesture-wise, but since I ultimately had to interact more with my own iPhone 5, you have no idea how many times I would try to swipe up from the bottom of the iPhone X to bring up Control Centre, only to either quit an application or go back to the Springboard. It was maddeningly frustrating. And when wanting to multitask, while I consciously processed the absence of a Home button in the iPhone X, I constantly had to pause for a fraction of a second, and recall the gesture in my mind. Much like when on the Mac you’re starting to forget a keyboard shortcut you don’t use that often.

Subjective experiences, however, shouldn’t be used to evaluate whether something is good design or not. The remark I’ve often heard from users who upgraded to X‑style iPhones since day one is that the newer interaction model ‘feels more natural’ than the older. I believe it’s an illusion. Which doesn’t mean I believe they’re lying. It means that before the iPhone X existed, the older interaction model felt just as natural as the new one today. This is what happens when your muscle or spatial memory is fully trained and accustomed to a way of doing things.

If we put apart ingrained habits, both interaction models have their merits, and neither is superior to the other as a whole.

• Double-clicking the Home button to bring up the multitasking interface is both quicker and precise than the mindful swipe from the bottom on X‑style iPhones. The gesture is unambiguous: double-clicking the Home button does only that; there is no overlap with another gesture.
• The gesture of switching between open apps appears to be more efficient in the X‑style iPhone interaction model, as opposed to always double-clicking the Home button and then leafing through the app ‘cards’.
• The gesture of invoking Control Centre is controversial. I can totally understand the dilemma here: a flick upwards from the bottom of the screen is by far the quickest gesture, therefore it should be associated with a high-priority/high-frequency task. In the absence of a Home button, multitasking and Control Centre became the two main contenders for that spot. I suppose a judgment call was made, and multitasking was deemed a higher priority. Quite understandable. But then where to move the Control Centre gesture? Swiping down from the top right corner feels like the kind of decision where someone in the room proposes it, there’s a long silence, and then they ask, Anyone has a better idea? And then there’s another long silence.
• As a result, the new X‑style iPhone interaction model has two gestures that partially overlap when you start from the bottom of the screen (quitting the current app, invoking the multitasking interface), and two gestures that partially overlap when you start from the top of the screen (invoking Notification Centre and Control Centre). In the older interaction model, these four gestures/operations are more distinctly assigned. You may consider a single click of the Home button to exit an app and the double click to invoke the multitasking interface as a partial overlap, but the two gestures don’t require precision or mindfulness on the user’s part to be consistently executed all the time. They’re clearly distinct and can be executed mechanically.
• Overall, the feature arrangement of the traditional iPhone interaction model seems to have a better degree of usability. Due to the increasing size of current flagship iPhones, the reachability of Control Centre in particular has unquestionably worsened.
• Using Apple Pay in the new interaction model is also a bit more cumbersome than it is on iPhones with a Home button. Where you simply placed your finger on the Home button sensor to authenticate and pay, now you have to double-click the side button, then glance at the iPhone to authenticate with Face ID.
• In theory, taking a screenshot has gone from a two-handed gesture on traditional iPhones (simultaneously press and then release the Home button and the Sleep/Wake button), to a potentially one-handed gesture on X‑style iPhones (simultaneously press and then release the side button and volume-up button). But this latter gesture makes you hold the phone awkwardly and you can accidentally press the volume-down button and end up performing the gesture to invoke the Emergency SOS.

I’d say that the X‑style iPhone interaction model is perhaps the best compromise they could come up with at Apple after removing the Home button. In the following gesture reshuffling, some gestures have turned out to be equally effective (exiting an app), and some even more effective (switching between open apps), but I don’t think we can conclude that it’s a better design as a whole. The two interaction models work well with the hardware they are implemented on, and work well when considered separately. They’re two different beasts. 

Making comparisons is where things get tricky and deeply subjective. As a user interface enthusiast, I’ve tried my best to point out some merits and downsides of the main gestures of each interaction model. Personally, I’m more accustomed to the Home button design, and it’s hard for me to view the X‑style iPhone interaction model as more than a necessary workaround, a gestural chain reaction following one single design decision — the removal of the Home button.