## Interstellar

I watched Interstellar last night. It’s rare that I don’t like any half-decent science fiction movie, so it gets a thumbs up, though it had its high- and low-points.

It would be difficult to get away with describing Interstellar as truly a “hard science” movie – but it makes quite a few nods in that direction, my favourite being its insistence that a worm hole, as an anomaly in three-dimensional space, should actually be a “worm sphere”.

The fundamental conceit of the film – that a hick farmer from the western US (or somewhere meant to look like the western US) was really a top quality pilot – was difficult to buy into while Michael Caine’s performance was universally dismal.

And, of course, the overall plot feels like an attempt to reimagine 2001: A Space Odyssey – which, despite being nearly 50 years old now, remains unsurpassed as filmic musing on humanity’s destiny in space.

## Pascal’s triangle and hypercubes

This blog entry poses a genuine question from me and, I hope, a puzzle that might interest many. It is inspired by an entry in the “Short Dictionary of Heuristic” found in How to Solve it: A New Aspect of Mathematical Method.

First, a quick description of Pascal’s Triangle.

This is the triangle of integers – with 1 at the apex – where each number is the sum of the numbers immediately above it – hence it begins (apologies for the slightly unorthodox presentation – a consequence of wordpress.com’s insistence on suppressing leading spaces on lines):

``` 1 1 1 1 2 1 1 3 3 1 1 4 6 4 1```

…and so on.

As is so often the case with maths, this is more than just a pleasing pattern. The numbers also, correspond, for instance, to the coefficients of the expansion of the binomial $(x+a)^n$ eg:

$(x+a)^0 = 1$
$(x+a)^1 = x+a$
$(x+a)^2 = x^2 + 2ax + a^2$
$(x+a)^3 = x^3 +3ax^2 +3a^2x + a^3$ and so on…

But it is the relationship of the triangle to geometry I am interested in.

The simplest (the only) null dimensional shape: a point, has 1 vertex: 1
The simplest 1 dimensional shape – a line: has 2 vertices (null dimensions) and 1 one-dimensional shape: 2 1
The simplest 2 dimensional shape – a triangle: 3 vertices, 3 one dimensional elements (bounding lines) and 1 two-dimensional space: 3 3 1
The simplest 3 dimensional shape – a tetrahedron: 4 vertices, 6 bounding lines, 4 two-dimensional elements (faces) and contains 1 three dimensionsal space: 4 6 4 1

So we can see Pascal’s triangle emerge there (and thus propose by induction that the simplest four dimensional space has 5 vertices, 10 bounding lines, 10 faces, contains 5 three dimensional spaces and 1 four dimensional space). This induction is correct – the shape is the pentachoron.

What puzzles me is whether we can draw any analogies for other shapes.

So for the families of squares and cubes:

We could have the same first two members – a point, 1
A line: 2, 1
Then a square: 4 vertices, 4 bounding lines, 1 space
A cube: 8 vertices, 12 bounding lines, 6 faces, 1 space
A hypercube (tesseract): 16 vertices, 32 bounding lines, 24 faces, 8 three dimensional spaces (cells) and 1 four dimensional space.

So we have a triangle like this:

```1 2 1 4 4 1 8 12 6 1 16 32 24 8 1```

My question is: is there some fundamental relation between the rows of this triangle and there is for Pascal’s triangle? I cannot see one, but there looks like there should be one!