“Describe yourself.” You can say it to people. And you can say it to numbers too. For example, here’s the number 3412 describing the positions of its own digits, starting at 1 and working upward:

3412 – the 1 is in the 3rd position, the 2 is in the 4th position, the 3 is in the 1st position, and the 4 is in the 2nd position.

In other words, the positions of the digits 1 to 4 of 3412 recreate its own digits:

3412 → (3,4,1,2) → 3412

The number 3412 describes itself – it’s autonomatic (from Greek *auto*, “self” + *onoma*, “name”). So are these numbers:

1

21

132

2143

52341

215634

7243651

68573142

321654798

More precisely, they’re __pan__autonomatic numbers, because they describe the positions of all their own digits (Greek *pan* or *panto*, “all”). But what if you use the positions of only, say, the 1s or the 3s in a number? In base ten, only one number describes itself like that: 1. But we’re not confined to base 10. In base 2, the positions of the 1s in 110 (= 6) are 1 and 10 (= 2). So 110 is __mon__autonomatic in binary (Greek *mono*, “single”). 10 is also monautonomatic in binary, if the digit being described is 0: it’s in 2nd position or position 10 in binary. These numbers are monoautonomatic in binary too:

110100 = 52 (digit = 1)

10100101111 = 1327 (d=0)

In 110100, the 1s are in 1st, 2nd and 4th position, or positions 1, 10, 100 in binary. In 10100101111, the 0s are in 2nd, 4th, 5th and 7th position, or positions 10, 100, 101, 111 in binary. Here are more monautonomatic numbers in other bases:

21011 in base 4 = 581 (digit = 1)

11122122 in base 3 = 3392 (d=2)

131011 in base 5 = 5131 (d=1)

2101112 in base 4 = 9302 (d=1)

11122122102 in base 3 = 91595 (d=2)

13101112 in base 5 = 128282 (d=1)

210111221 in base 4 = 148841 (d=1)

For example, in 131011 the 1s are in 1st, 3rd, 5th and 6th position, or positions 1, 3, 10 and 11 in quinary. But these numbers run out quickly and the only monautonomatic number in bases 6 and higher is 1. However, there are infinitely long monoautonomatic integer sequences in all bases. For example, in binary this sequence at the Online Encyclopedia of Integer Sequences describes itself using the positions of its 1s:

A167502: 1, 10, 100, 111, 1000, 1001, 1010, 1110, 10001, 10010, 10100, 10110, 10111, 11000, 11010, 11110, 11111, 100010, 100100, 100110, 101001, 101011, 101100, 101110, 110000, 110001, 110010, 110011, 110100, 111000, 111001, 111011, 111101, 11111, …

In base 10, it looks like this:

A167500: 1, 2, 4, 7, 8, 9, 10, 14, 17, 18, 20, 22, 23, 24, 26, 30, 31, 34, 36, 38, 41, 43, 44, 46, 48, 49, 50, 51, 52, 56, 57, 59, 61, 62, 63, 64, 66, 67, 68, 69, 70, 71, 75, 77, 80, 83, 86, 87, 89, 91, 94, 95, 97, 99, 100, 101, 103, 104, 107, 109, 110, 111, 113, 114, 119, 120, 124, … (see A287515 for a similar sequence using 0s)

In any base, you can find some sequence of integers describing the positions of any of the digits in that base – for example, the 1s or the 7s. But the numbers in the sequence get very large very quickly in higher bases. For example, here are some opening sequences for the digits 0 to 9 in base 10:

3, 10, 1111110, … (d=0)

1, 3, 10, 200001, … (d=1)

3, 12, 100000002, … (d=2)

2, 3, 30, 10000000000000000000000003, … (d=3)

2, 4, 14, 1000000004, … (d=4)

2, 5, 105, … (d=5)

2, 6, 1006, … (d=6)

2, 7, 10007, … (d=7)

2, 8, 100008, … (d=8)

2, 9, 1000009, … (d=9)

In the sequence for d=0, the first 0 is in the 3rd position, the second 0 is in the 10th position, and the third 0 is in the 1111110th position. That’s why I’ve haven’t written the next number – it’s 1,111,100 digits long (= 1111110 – 10). But it’s theoretically possible to write the number. In the sequence for d=3, the next number is utterly impossible to write, because it’s 9,999,999,999,999,999,999,999,973 digits long (= 10000000000000000000000003 – 30). In the sequence for d=5, the next number is this:

1000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000005 (100 digits long = 105 – 5).

And in fact there are an infinite number of such sequences for any digit in any base – except for d=1 in binary. Why is binary different? Because 1 is the only digit that can start a number in that base. With 0, you can invent a sequence starting like this:

111, 1110, 1111110, …

Or like this:

1111, 11111111110, …

Or like this:

11111, 1111111111111111111111111111110, …

And so on. But with 1, there’s no room for manoeuvre.