3 ms·
That raises the question: how much faster can we read without speaking, than read aloud? I'm guessing its a number less than 2 times.
by kristianp 2y ago
That raises the question: how much faster can we read without speaking, than read aloud? I'm guessing its a number less than 2 times.
- soperj 2y agoI can definitely read 4 times faster in my head than out loud. I've read books to my kid out loud that take like 30 minutes per chapter, and i doubt it would take more than an hour to read the whole book. I read on the faster side though, and don't have an internal monologue.
- nicce 2y agoI can also read about the same speed but I don’t enjoy reading at that speed at all...
- ta2112 2y agoFor me it's 1:1. I read aloud in my head at the same pace that I speak. It seems slow, but little by little I get through what I need. Also, I don't know how to do it another way. So it goes.
- stevenwoo 2y agoIt depends upon the material for me. Sometimes with technical or historical books that are essentially reciting data points then conclusions I can read much faster than 2x because if one is paying attention one kind of knows where the writer(s) are going - though I might repeat or pause to write notes or do online research sometimes to memorize and understand new terminology, but for fiction I have to be in a very specific mindset to imagine all the things going on appropriately with movement and emotion and pondering character motivation and the setting, sometimes I just pause and think about what is being described in my minds eye before moving forward in the text.
- aidenn0 2y agoMy wife asked me this question, since I'm a considerably faster reader than her. We discovered that (if I try reading aloud as fast as possible) I can read aloud faster than she can read without speaking, though my tongue does occasionally trip over words at that speed. I read in my head faster yet.
- alwa 2y agoOh man. I’d go higher. Especially compared to vocalizing at “storytelling pace.” I remember a while back being surprised to learn that there are readers who process language phonographically as opposed to orthographically, even when silent-reading. That is, they might read in their minds at the same pace that they speak, or “subvocalize”—reading aloud in everything except for producing the sounds. It was a dinner-party-grade conversation, not an academic-grade one, but once somebody talked to me about that distinction I’d notice people mouthing along to what they were reading—and it made sense why that was slower. I’m not up on the details, but I remember an estimate in the 150-200wpm range for reading out loud. A little north of that for silent reading with subvocalization, and comfortably in the 400-600wpm range for skilled silent readers in an orthographic mode. And then the whole “speed reading” crowd clocking in faster, but at significant cost to comprehension. The general idea being that skilled orthographic readers can “chunk” entire words and phrases into larger visual units that they can parse all at once. Looking for references now this is the best I can find: https://theamericanscholar.org/reading-fast-and-slow/ https://theamericanscholar.org/reading-fast-and-slow/ But I wonder where the speed reading conversation intersects with this notion out of Caltech that we “live at 10 bits per second”: https://news.ycombinator.com/item?id=42449602 https://news.ycombinator.com/item?id=42449602