Speedreading x100: from vocalization to sub-visualization

What if I told you that you can improve your reading speed not by x2 or x3 but x100 without losing comprehension? This is speedreading on steroids! Going from vocalization (100 wpm) to sub-visualization (10 000 wpm) is just that. You might think I am crazy. Yet this is exactly what happened to me. Everybody can do this. Here I want to describe some of the main milestones along the path.

TLDR:

See my masterclass

https://keytostudy.thinkific.com/courses/keytostudy-speedreading-masterclass-1

The masterclass is quite long and dense, like 14 mini-courses.

Practice for a decade and you will read faster than I do.

Now you get the story.

Starting point

I used to read relatively slowly and my retention was not stellar, especially when I was stressed or did not like what I read. When Anna first tested me, he openly laughed and mocked me. Then we worked for several months until my retention was near 100%. The reading speed actually dropped to 100 words per minute. This is my vocalization speed. This happens a lot – because we visualize everything. Practicing memory and retention we remember more details but lose some of the speed.

It is important to say that at this point I was mainly reading nonfiction books with paper covers, and articles printed on A4 paper. Most of the stuff I used for speedreading practice dealt with business and history.

Speeding up

Going from 100wpm to 400wpm is relatively straightforward.

• I dropped the details from the visualizations I did not need. The visualizations became plain and almost icon-like.
• Instead of complex mental palaces with 300 locations per palace, I switched to mindmaps. Mindmaps follow better the text structure, like a table of contents.
• The retention dropped from near 100% to 85% which was still more than acceptable.
• I used metaguiding, moving the page down 10% faster than I could comfortably read.

The improvement was fast. In my case, it happened within a month. And it is also safe.

Subvocalization suppression

Going above 400 wpm requires so-called subvocalization suppression. Subvocalization is the inner voice we hear as we read. Our talking speed is limited, and we want to remove this limitation.

I waited to have at least 85% retention using visual mental structures to do it. Then I counted while reading 21-22-23. The idea is to understand the text from the visual cues only. At first, I understood nothing. After 20 min I started understanding something. After two weeks I was able to read 400 wpm with 85% retention while counting. I continued to do that until Anna suggested stopping counting. I did not count anymore, but the inner voice was gone.

It is OK to subvocalize the names and the places we encounter for the first time, and maybe small percentages from the words within the text.

Saccading

The next part of training was almost painful. I had to use a lot of Pomodoro breaks. I think 60% of my training time at this stage only I sat with my eyes closed so that they will not hurt.

The eyes kind of naturally glide along each line. Instead of gliding, saccading assumes jumping the focus through three columns of the text. To do that, the visual angle should be wide enough to cover an entire column. It took me around 1 month of training with tablesto get there. Then all I had to do was move my eyes faster. Again, metaguiding with a page does this trick. I got from 400wpm to 800 wpm without losing retention.

Approximately at this point the 1:1 with Anna ends. I stopped working with Anna and focused on my own needs.

Adding complications and strategy

I used the next 4 months to train with other materials. My original practice was in English. Now, I practiced 2 more languages until I got full comprehension. Then I added a large range of materials: math, code, online articles, and poetry. Immediately I understood that I need to adapt my reading strategy to the material that I read. Sometimes, I could skip prereading, other times I needed to reread many times, and in other cases, I had to read at vocalization speed.

Adding a flexible reading strategy is not a trivial step. It requires a lot of trial and error. At this point, I switched from printed books to online blogs and articles, like techcrunch and business insider.

I was reading well, but my eyes were still tired. The Pomodoro breaks at this stage constituted 30% of my reading time.

Using eidetic memory

I noticed that instead of reading all saccades of all columns I could read diagonally, moving line by line as well as saccading left to right. This mechanism uses so-called eidetic or photographic memory. There are stories of individuals with stable photographic memories. I am not one of those people. My eidetic memory lasted for 3 lines for 1 second. And yet my saccading reduced x3. This was a blessing for my eyes. Pomodoro breaks got back to normal 10 min for every hour. There was no effect on reading speed or retention.

To be honest, I practiced eidetic memory from day one of training: scanning images with my eyes and then recollecting what I saw from my memory. I was not sure why this exercise was good, until the point I needed eidetic memory for reading.

Another thing I continued doing was working on tables for the visual angle. I did not stop at 3 columns, and with time I could read 12 words by 5 lines in one glance. It is important to notice that it took me well more than a year of daily practice to do this.

Visual flow is the key to real speedreading

The next step of speeding up is a bit like magic. I focused in the middle of the paragraph and closed my eyes. Then I read everything I got in peripheral vision from my memory. Then I moved my eyes 5 lines down. This is a temporary step.

Next, I stopped closing my eyes and simply moved them from top to bottom of a computer screen in a portait mode. I suggest by this stage to set up one of your computer screens in portrait mode and read from it. At high speed, even turning pages of a paper book is a pain. This method is better.

Suddenly my reading speed went to 1600 wpm. but my retention dropped significantly. I could use visual flow only with the stuff I did not really care about. So I read significantly more history. Not very dense texts. And yet my comprehension was not fast enough.

Speedreading at 1600 wpm is something that my students almost never report. I think the main limitation is not speedreading technique itself, but the choice of materials. There is a very clear separation between the content we need for actual progress and the content used specifically for speedreading training. At high reading speeds, it is natural to forget that.

Adding massive memory structures

My progress in retention at 1600wpm was due to using a new set of memory structures and reading in a very specific way.

1. Divide the text into sections. Each section is usually a long paragraph of 7 lines or so. If the paragraph is very long or very dense it can have several sections. If the text is not very informative, a section can be two or three pages.
2. Per section chunk between 5 and 12 words. A 3×3 words matrix usually does this trick. Later I used a sparse 5×5 matrix. The words are selected and chunked while reading. The next part- the visualization – is performed while saccading (eyes unfocused, residual image gone).
3. The lines in the chunking matrix become person-action-object of the relevant PAO visualization.
4. The PAO is put in a modified mental palace. At this point, I used so-called Vaughn cube model which I learned from Anthony Metivier and a very simple itinerary in the mental palace.  The idea is to put 1 PAO in each corner of the palace and add metadata on the floor and ceiling. The use of the walls is contextual: PAO, flowcharts, graphs, or something else.
5. Per page, chapter, or some other large section, go to the next room of the palace. The color of the room is determined by the subject discussed: white for info, green for innovation, blue for methodology, yellow for actionable items, and red for motivational stuff.

I also added revisiting the palace upon completion of a small article or a chapter of the larger paragraph. During revisited I entered into the visualized rooms reviewing the visualization, but I also added flowcharts with the logical conclusions on the walls of the palace, and associations with other stuff. I used to ask questions and play “what if” games. While the reading speed was 1600wpm when reading, I spent an equal amount of time revisiting the mental structures and analyzing the content. Due to the revisiting stage, I did not really need Pomodoro break anymore. My short-term retention went up, but I did not remember much the next day.

Developing speedwriting

At this point, most of my reading was dedicated to supporting my writing. I was working on my blog and memory course.  I could no longer use spaced repetition. So I started to use a simple reading diary, with 5 keywords per article – the anchor marker for the relevant memory palace. And yes, my knowledge base became sufficiently complex to use elaborate mental landscapes instead of simple mental palaces of mental streets.

Speedreading naturally requires notetaking so we can see which memory structure to revisit. Yet, with notetaking, it is really easy to create content. Speedwriting reduces the need to revisit memory structures. I could read much more, and partially due to notetaking, I could retain the information for years.

Speedreading on steroids: from 1600wpm to 10000wpm

Now the jump from 1600wpm to 10000 wpm was very abrupt. Somehow I noticed that I do not visualize anymore. Just like kids switch from vocalization to subvocalization, I switched from visualization to subvisualization.

Immediately my retention went down. I could use 10000 wpm speed only for prereading: choosing which parts of which article need detailed attention. Then I reread with 1600wpm speed, but only the relevant parts. The effective reading speed was 3000wpm.

The magic of subvisualization happened when I was working on context switching for productivity. At this point, I was reading at work, while the code was running, and I constantly needed to switch focus between the search window, the reading window, the code execution window, and the visualization of the result. This required context switching. It somehow worked, but I do not understand the mechanism well enough.

Innovation focus: improving retention at 10000wpm

With time, I started to read much more at 10000wpm. That was my speedreading goal all along. Basically, rereading got very limited. This happened due to the innovation focus. At this time I was using visual dictionaries, e.g. reusing the same visualization. I read roughly about the same subjects for years. I revisited the memories and updated reading notes for 2 sec per article because the articles did not really contribute – except for small innovation. So to simply got to encode very little information per article.

The first overview article I read is usually 1600wpm. By the 10th article, I average 3000wpm. By the 40th article on the same paradigm, I read 10000 wpm, because it adds so little new information. And my retention does not drop. unless I forget to change my reading strategy.

Speedreading requires at least 50% comprehension. I get 70%. It is not my fault that 90% of that number comes from my prior knowledge. The innovation is precious and rare/

Conclusion

I read a lot on different subjects in different languages – almost always from a second screen. Either the screen is in portrait mode, or I arrange the window to take 40% of the screen in landscape mode. I read a lot for more than a decade, and I always think about my practice – not when reading or analyzing but between tasks. I put my methodology in a simple masterclass.

https://keytostudy.thinkific.com/courses/keytostudy-speedreading-masterclass-1

The masterclass is quite long and dense, like 14 mini-courses.  It took me a decade to get to the reading speed I needed. But I am not a very talented memory master and my reflexes are naturally slow. Take my course, practice for a decade and you will read faster than I do. Maybe even 30, 000 wpm.

[There is a time-limited 50% discount coupon keytostudy_memory_speedread_50, and it applies to the recommended bundle

https://keytostudy.thinkific.com/bundles/keytostudy-triple-memory-triple-reading-speed-triple-focus

as well as its components

https://keytostudy.thinkific.com/courses/keytostudy-memory-masterclass,

https://keytostudy.thinkific.com/courses/keytostudy-speedreading-masterclass-1,]