Progressive deepening

Concepts by the slice

When humans have complex tasks, they break them into pieces, then do one piece at a time. It’s the same with learning something complex. Break it into topics, then learn one topic after another. The topics are broken into subtopics, and so on, until you get bite-sized pieces that your brains can handle.

Often there’s more than one way slice content. When you write a Cycourse, you – the author – control how that slicing is done. For example, let’s say you’re writing a course on the glorious country of Pottsylvania. You could slice it by:

  • History
    1. 4,000 BCE – 500 CE: Tribal settlement and expansion
    2. 500 CE – 700 CE: The British Invasion
    3. 700 CE – 1,500 CE: The Great Snooze
  • Economics
    1. Agriculture
    2. Mining
    3. Manufacturing
  • Politics
    1. National government
    2. The central cantons
    3. Cantons of the boonies …

How you slice your content is a Big Deal. How experts slice content might not be good for novices.

Slicing by technology

An example. Let’s say you’re going to make a Cycourse on how to make Web pages. The course will help students learn two technologies: HTML and CSS. (Don’t worry if you don’t know what these things are.)

One way to slice the content is by technology. Here are the chapters from a book that you can buy:

  1. Structure
  2. Text
  3. Lists
  4. Links
  5. Images
  6. Tables
  7. Forms
  8. Extra markup
  9. Flash, Video & Audio
  10. Introducing CSS
  11. Color
  12. Text
  13. Boxes
  14. Lists, Tables & Forms
  15. Layout
  16. Images
  17. HTML5 Layout
  18. Process & Design
  19. Practical Information

The first nine chapters are about HTML. The next eight are about CSS. That’s slicing by technology. The last two chapters are about the process of using the tech.

In the HTML part, the tags get more complex. Text tags are easier than list tags are easier than link tags, and so on. Again, slicing technology.

It’s the same in the CSS part. Color properties are easier than text properties, etc. Layout is the hardest of all.

There’s a problem here. In RL (real life), an HTML page will have text, links, images, etc. HTML and CSS are used together as well. RL doesn’t slice by tech, the way this book does.

Every book and Cycourse should have lots of examples. If you slice by tech, you may have to use artificial examples. You write about tables and forms without using CSS, since CSS is later in the book. In RL, people don’t use tables and forms without using CSS.


You lose context. Context means the goals and constraints of a task. In RL, context determines how you do tasks. One thing that makes experts experts is that context drives their choice of tech, not the other way around. Students need to learn why a task is done one way in one context, and another way in another context.

Context helps motivate students. They’re not just learning tech, they’re learning how to do useful things.

Progressive deepening

There’s more than one way to slice a Cycourse. Here’s an opening chapter from the venerable CoreDogs (with some stuff omitted):

  1. A Web page with text
    1. The structure of a Web page
    2. Basic HTML tags
    3. Simple font tags
    4. Adding style with CSS
    5. Making lists
    6. Writing for the Web

The chapter is about a task: making Web pages with text. That’s the context. The chapter introduces a few HTML tags that are useful when making text pages. Then it adds a little CSS that is useful for the same tasks. Then list tags. Not everything about them. Just enough to get started.

The last topic isn’t about tech at all. It’s tips for writing, like “Use active voice” and “Write for scanning.” In the tech-sliced book we talked about, process advice like this was an afterthought. In CoreDogs, it’s attached to the task contexts where it’s relevant.

That’s the first slice. A simple task, then all (well, some, anyway) of the pieces needed to do that task.

The second slice has another task. It adds a few more pieces of tech, like links and images. It expands on the tech introduced in the first slice, explaining more about text and list tags, and their CSS. It adds some workflow stuff as well, like how to group images and captions so readers see them as belonging together.

The third slice has yet another task, or more complex versions of the tasks already used. This slice introduces layout, and more about links and images. More workflow, e.g., how layouts can include reusable content, to make site maintenance easier.

This way of slicing is called progressive deepening. Slice by context, not by tech. A simple task with a simple context, covering the tech needed to do the task. Another task with another context, adding a bit more tech, and so on. The tasks are chosen to introduce more and more tech.

There’s a difference in course outcomes. In both the tech-sliced and context-sliced courses, students learn the same HTML and CSS. In the context-sliced course, they also learn how to think about tasks.

Visualizing the difference

Here’s a visualization of the difference between tech slicing and context slicing. Let’s start with the topics to be covered:


Here’s the tech-oriented way to slice the content into chapters:

Tech oriented slicing

Each of the blue boxes is a chapter.

Here’s the task-oriented way to slice it:

Task oriented slicing

The first chapter has a task that needs simple HTML and simple CSS. The next chapter’s task adds some less simple HTML and CSS, and so on.

Topic slicing is horizontal. Context is secondary.

Context slicing is vertical. Context is always there.

Repeating content in different contexts

Progressive deepening has an interesting advantage: each chapter would necessarily include a new version of the stuff covered in the previous chapter. That is, chapter two would have a review of chapter one concepts in a different context, before it starts talking about the less simple HTML and CSS. By the end of the course, students might have seen the simple HTML in five different contexts. The explanation of the basics could get less and less each time. The last couple of chapters might just give the simple HTML and CSS, without explanation.

You could use the same context throughout the course. Part of it at the beginning, then more of it, and so on. That’s efficient. However, students wouldn’t see contextual differences, learn what generalizes across contexts, and what doesn’t.

Of course, there’s no such thing as free time. Time spent learning contexts would not be spent learning tech. Is the tradeoff worthwhile? How much of a context should be explained, before the details get overwhelming? That’s for you to decide.


When humans have complex tasks, they break them into pieces. It’s the same with Cycourses. Suppose you’re writing a Cycourse on Web tech. One way to slice the content is by technology. You lose context. Context means the goals and constraints of a task.

Progressive deepening means to slice by context, not by tech. A simple task with a simple context, covering the tech needed to do the task. Another task with another context, adding a bit more tech, and so on.


Add new comment

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Prove that you are sentient.
1 + 0 =
Solve this simple math problem and enter the result. E.g. for 1+3, enter 4.