How to keep a long conversation readable on a phone screen

You are sitting in a long meeting, an academic lecture, or a multi-speaker discussion, trying to follow along by reading a running transcript on your mobile phone. After a few minutes, the words start to blur. Your eyes feel dry, and you realise you have read the same line three times without taking in its meaning.

Reading live speech as it is spoken is fundamentally different from reading a static book or a pre-edited article. A live transcript is dynamic, unpredictable, and visually unstable. When you read live text for more than a few minutes, your working memory and visual system are put under immense strain.

To make this experience sustainable, we have to look at the ergonomics of reading. By understanding how the human eye processes text on screen, we can configure our devices and our reading habits to prevent cognitive fatigue.

The visual mechanics of reading on a small screen

When we read, our eyes do not move in a smooth, continuous sweep. Instead, they make rapid, micro-jumps called saccades, interspersed with brief pauses called fixations. It is during these fixations that the brain actually processes the letters and extracts meaning.

On digital screens, this mechanical process is highly sensitive to text complexity. An eye-tracking study published in the journal of the Association for Research in Vision and Ophthalmology found that complex or poorly formatted text significantly increases both the duration of each fixation and the number of regressive saccades—the backward jumps your eyes must make to re-read text when comprehension fails.

This physical strain is compounded by how we naturally scan digital content. When presented with dense blocks of text, our eyes tend to fall into an "F-pattern". As detailed by the Nielsen Norman Group's eyetracking research on scanning patterns, readers scan horizontally across the top lines, drop down slightly for a shorter horizontal scan, and then skim straight down the left margin.

If a live transcript is presented as an unbroken, continuous block of scrolling text, your eyes will inevitably drop into this F-pattern. You will read the first few words of each new utterance but miss the critical context at the end of the sentence. To prevent this, the interface must break the text up to disrupt the F-pattern and encourage deep, sustained reading.

Designing typography for the scrolling page

The most effective way to prevent skimming and reduce eye strain is to control the visual structure of the text. This starts with the horizontal boundaries of the screen.

In print and desktop publishing, the standard recommended line length is between 50 and 75 characters, according to the Baymard Institute's readability studies. However, on the narrow viewport of a mobile device, this width forces the eye to make rapid, exhausting sweeps. To keep a readable live transcript phone screen layout, you must shorten the visual width of the text. For live speech captioning, the University of California's captioning guidelines recommend limiting lines to 32 to 42 characters, with no more than two or three lines of text displayed per block.

Vertical density matters just as much. Large, academic-sized paragraphs of 150 words do not work on a phone. Instead, speech should be broken into scannable paragraphs of two to four sentences. These short blocks introduce essential whitespace, giving your eyes a natural point of rest between speakers and ideas.

Controlling motion and stabilizing live text

A static document stays where you put it, but a live transcript is constantly in motion. Words appear at the bottom, shifting older text upward. Even worse, automatic speech recognition systems often correct previous words retroactively as they gain context, causing the text to jump and flicker.

This visual instability is a major driver of reading fatigue. Research published by CHI on modelling and improving text stability in live captions shows a direct correlation between text flickering and elevated viewer distraction. When words jump unexpectedly, your eye loses its fixation point and must perform corrective regressive saccades to find its place again.

To make a continuous transcript usable, you must have control over this motion. Standard accessibility principles dictate that any auto-updating or scrolling text must feature a mechanism that allows the user to pause, stop, or hide the movement.

For a system-wide utility, the iPhone has a built-in feature called Live Captions, which converts ambient spoken audio into a floating text overlay. This works well for brief alerts or casual interactions. However, for a structured, long-form discussion, you can use the Hontoni app to view a formatted transcript or a written translation. In Hontoni, you tap the Listen button to start the microphone, and the text arrives with clear speaker labels, allowing you to scroll back to review earlier parts of the conversation.

Understanding and managing cognitive language fatigue

Reading a continuous stream of live speech is cognitively demanding because your brain is trying to synthesise visual and auditory information simultaneously. When the material is dense, or when you are reading a translation, this processing can lead to a state of exhaustion.

This is known as cognitive language fatigue. As explained in Lissie Albornoz's analysis of cognitive language fatigue, when your working memory becomes completely overloaded, your brain begins to shut down non-essential processing. You might find yourself staring at the words without registering their meaning, experiencing a physical "mental wall."

To manage this fatigue during a long conversation, you must build cognitive pauses into your reading session:

• Use the pause feature: Do not let the continuous scroll dictate your pace. Pause the transcript feed to read a block of text at your own speed, then jump back to the live edge when your brain has processed the information.
• Set a strict session limit: Cognitive endurance for focused reading typically peaks at 20 to 30 minutes. Schedule a physical screen break every half hour to look away, rest your eyes, and let your working memory clear.
• Recognise the technical limits: Automated transcripts are highly helpful, but they make mistakes. Background noise, strong accents, and people talking over one another will increase the error rate, which in turn increases your cognitive load as you try to decipher incorrect words. Accept that some sections will be imperfect, and do not waste mental energy trying to parse unreadable blocks.

A practical step for your next session

Keeping a running transcript readable is about pacing both the screen and your mind.

Before your next long meeting or lecture, decide on a visual strategy. Rather than holding your phone and watching the text scroll continuously, place the phone on a stable surface at a comfortable viewing distance. Set a timer on your watch for 25 minutes. When the timer goes off, pause the visual flow of the transcript, look away from the screen for two minutes to let your eyes recover, and then return to the text.

If you need to follow an extended discussion in another language, Hontoni may suit your situation. The app displays a live transcript or a live written translation on your screen as people speak, which can help you stay engaged without hitting a mental wall. However, because automated translation struggles when multiple people speak at once in noisy rooms, you should test it in your typical environment first. For high-stakes medical or legal settings, a professional human interpreter remains essential.