1
00:00:14,440 --> 00:00:20,560
So I want to welcome you to the Mathematical Institute for Not a Public Lecture.
2
00:00:21,040 --> 00:00:22,630
So this one is very special.
3
00:00:23,110 --> 00:00:34,850
Martin Bryson started as head of department this October and we thought it would be a good idea for him to, to, to be to give a public lecture.
4
00:00:34,870 --> 00:00:41,860
I see is the public face of the department and to present him to the whole public as well as the staff.
5
00:00:43,570 --> 00:00:50,799
It's a very special one because it's an inaugural lecture, but is the first inaugural lecture for head of department in this department.
6
00:00:50,800 --> 00:00:54,640
So it's the inaugural inaugural lecture of the department.
7
00:00:55,540 --> 00:01:03,670
So before I introduce Martin, let me remind you that in case of emergency, we have exits here and there.
8
00:01:04,150 --> 00:01:08,260
If you go through this door, you'll go down and then you find the exit.
9
00:01:08,260 --> 00:01:16,899
But make sure not to disturb the ghost of all the head of departments, previous heads of department still roaming around into the never.
10
00:01:16,900 --> 00:01:20,470
They never quite leave. Okay.
11
00:01:22,150 --> 00:01:27,070
I actually wanted somehow we sung the previous head of the department, Professor Harrison,
12
00:01:27,070 --> 00:01:32,560
to give the introductory remark so they could do all the usual jokes about being head of department.
13
00:01:32,890 --> 00:01:41,290
But he was very wise. He left the country right after he was done, but he left a couple of notes for Martin.
14
00:01:41,560 --> 00:01:45,790
So I hear the first envelope. So let's see.
15
00:01:47,650 --> 00:01:52,180
It's quite short. Dear Martin, there is no money left.
16
00:01:58,170 --> 00:02:03,540
Best wishes and good luck. Okay, so that settles it.
17
00:02:03,540 --> 00:02:06,360
So now we know why he's not here anymore.
18
00:02:07,590 --> 00:02:17,370
So the second one with letter number two and number two says to be open only in the case of a massive, massive invasion of Belgium mathematicians.
19
00:02:18,330 --> 00:02:21,809
So I don't think that would happen. Okay.
20
00:02:21,810 --> 00:02:26,550
So let me tell you a little bit about about about Martin.
21
00:02:27,330 --> 00:02:32,490
Martin did his undergraduate here out for college in the Mathematical Institute.
22
00:02:32,760 --> 00:02:40,110
After that, he left. And it is a Ph.D. in Cornell in New York State and worked for a while.
23
00:02:40,110 --> 00:02:45,390
Professor at Princeton University, before he joined Oxford in the nineties,
24
00:02:45,660 --> 00:02:50,730
is currently the white head professor of pure mathematics at the University and a fellow of modern college.
25
00:02:51,300 --> 00:02:57,780
Its main interest are in geometry, topology and mostly in geometry group theory.
26
00:02:57,780 --> 00:03:03,360
So that's a lot of work and I hope that today's lecture will go into detail and explain that to us.
27
00:03:03,630 --> 00:03:07,770
So thank you very much, Martin, for accepting to give this lecture.
28
00:03:16,050 --> 00:03:22,650
And I thank you all for that very nice invitation. So since there's no money left for myself, just concentrate on the mathematics.
29
00:03:24,990 --> 00:03:28,740
So we're giving public lectures.
30
00:03:29,130 --> 00:03:33,959
I much prefer to do it in a village hall or a school where you're sure the audience,
31
00:03:33,960 --> 00:03:37,050
the trouble of giving it in the maths department is that the mathematicians show up.
32
00:03:39,090 --> 00:03:41,850
And so I have to say, I'm not going to talk to you, my colleagues.
33
00:03:41,850 --> 00:03:49,500
I'm going to concentrate on the the genuine public, as it were, and the genuine public, particularly the young people, I would say.
34
00:03:50,910 --> 00:03:58,080
I'm going to flash a lot of things at you. And don't worry if the details pass you by, the bound to pass you by.
35
00:03:58,080 --> 00:04:06,690
There's far too much information on these slides to absorb in a sort of symbol by symbol way or even a line by line way.
36
00:04:06,840 --> 00:04:11,500
Just let it wash over you and and feel the beauty of the thing.
37
00:04:11,570 --> 00:04:19,740
I just. Just try and hang on to the big ideas and let the pictures tell you something that you don't have to translate into words or symbols.
38
00:04:20,940 --> 00:04:24,450
So I'm going to talk about the words in the title. Very efficient of me.
39
00:04:24,930 --> 00:04:26,370
I'm going to talk about symmetry.
40
00:04:26,770 --> 00:04:33,180
I'm going to talk about spaces, that is to say, geometry and different sorts of environments in which geometry takes place.
41
00:04:33,780 --> 00:04:40,049
And in particular, as I said, I'm interested in geometric group theory or geometric group theory.
42
00:04:40,050 --> 00:04:46,320
A lot of it's about is how you can extract geometry from algebraic descriptions of symmetry.
43
00:04:46,320 --> 00:04:54,360
So I'll try and give you some sense of that, the way in which geometry emerges from algebraic structures that are designed to describe symmetry.
44
00:04:55,140 --> 00:05:03,210
And then there's this word undesired ability that my spellcheck keeps telling me doesn't exist but is commonly used and in mathematics,
45
00:05:03,450 --> 00:05:16,140
and that I want to try and convey to you that many natural problems in mathematics turn out to be too hard to expect as a global algorithmic solution.
46
00:05:16,230 --> 00:05:18,180
Okay. And so we'll see examples of that.
47
00:05:18,390 --> 00:05:24,720
And I'll try and convince you that this isn't some fringe phenomenon way beyond the boundaries of mathematics,
48
00:05:24,900 --> 00:05:30,690
mathematics fading into the sort of logic that only interests old men in wheelchairs with long beards,
49
00:05:30,870 --> 00:05:34,110
but it's really close at hand in natural problems.
50
00:05:34,350 --> 00:05:40,050
So that's why I want to talk about symmetry, the way that geometry emerges from thinking about symmetry,
51
00:05:40,410 --> 00:05:48,210
and then we'll try and touch on on the side ability and how it really comes, comes into play with with core mathematical problems.
52
00:05:49,530 --> 00:05:56,160
Right? So let's start with something we think we know. So we think we live on the surface of a sphere, that particular sphere.
53
00:05:56,610 --> 00:06:01,350
Okay, we might question that a little bit later, but let's think that's home, first of all.
54
00:06:01,860 --> 00:06:04,920
Now we might zoom in a bit.
55
00:06:06,260 --> 00:06:14,060
On those European islands there. If you don't know me, you might be puzzling as to my accent and exactly which one of those islands I'm from.
56
00:06:15,140 --> 00:06:19,340
We won't take a vote. I'm from that one in the middle.
57
00:06:20,210 --> 00:06:24,510
It's the Isle of Man. If you don't know me, you probably didn't that.
58
00:06:25,010 --> 00:06:30,440
And now the flag of the Isle of Man is what's called the Tri Scullion.
59
00:06:31,250 --> 00:06:37,340
That Viking symbol up there. And the motto that goes with it is Coca-Cola to stop it.
60
00:06:37,550 --> 00:06:45,150
Whichever way you throw me, I stand. Now, that's rather a nice symbol, and we can talk about symmetry.
61
00:06:45,170 --> 00:06:49,700
So let's start with that. So what symmetries does it have and what is a symmetry?
62
00:06:50,330 --> 00:06:59,360
Well, a symmetry is where you take something and then you say to the person observing you, close your eyes, you do something to that thing.
63
00:07:00,050 --> 00:07:04,280
And they say, Open your eyes, and they can't tell whether you did anything or not. That's what a symmetry is.
64
00:07:04,820 --> 00:07:10,830
You transform the object, whatever the object may be, in a way that doesn't destroy any of its essence.
65
00:07:10,850 --> 00:07:18,290
That really leaves everything a central preserved. So here I took this picture.
66
00:07:18,860 --> 00:07:22,190
I left alone, and I took this picture and I rotated it.
67
00:07:23,080 --> 00:07:29,229
And I took this picture and I rotated it twice. If you know how weak my computer skills are, you won't believe me.
68
00:07:29,230 --> 00:07:34,150
You'll know. I just cut and paste but just pretends. But that's what I did.
69
00:07:34,150 --> 00:07:39,969
It would have the same effect. So not so called rotation.
70
00:07:39,970 --> 00:07:47,500
Rotate by one click. And I've just noticed here that if you do it twice, that's the same as if you did the inverse operation of the first click.
71
00:07:47,500 --> 00:07:54,430
That is, instead of clicking to the to the right, I might have to anticlockwise I might have clicked one clockwise.
72
00:07:54,760 --> 00:08:03,909
Okay. Okay that this little symmetric nice little bit of symmetry and when I what let's think so that has three fold symmetry right.
73
00:08:03,910 --> 00:08:08,980
Has three symmetries. Do nothing, rotate once or rotate twice now.
74
00:08:09,280 --> 00:08:12,380
So that motto come quite a stab at that.
75
00:08:12,400 --> 00:08:15,910
That means whichever way you throw me, I stand in English.
76
00:08:16,720 --> 00:08:20,890
Now, in pigeon mathematics, we might say what we've just observed.
77
00:08:21,100 --> 00:08:27,910
It is, I have three symmetries in more sophisticated, fluent mathematics.
78
00:08:28,270 --> 00:08:35,889
It is something about my symmetry group statement, not just about how many symmetries I've got, but what those symmetries are.
79
00:08:35,890 --> 00:08:39,790
And so I might sort of be being a studious person.
80
00:08:39,790 --> 00:08:44,920
Write down the symmetries I found. I found I could do nothing, which I write as one for doing nothing.
81
00:08:45,280 --> 00:08:47,679
I could have rotated at one, so I could have rotates it twice.
82
00:08:47,680 --> 00:08:55,390
And that's all okay, now in my notebook that I'm going to keep because I'm going to try and make an extensive study of symmetry.
83
00:08:55,750 --> 00:09:00,700
I'm going to write more brief notes. I'm to say, Well, there's only one serious thing you can do to this.
84
00:09:00,700 --> 00:09:07,059
You can rotate it, and I'll make a note for myself that I don't forget in case I lose my flag,
85
00:09:07,060 --> 00:09:11,500
or at least to my notes, that if I rotate three times, I get back to where I started from.
86
00:09:11,720 --> 00:09:15,730
Rotating three times is the same as doing nothing. Okay, everyone happy?
87
00:09:15,910 --> 00:09:21,520
So just make some notes. Right now, mathematicians should be feeling a warm glow at this point.
88
00:09:21,520 --> 00:09:25,390
We're doing mathematics. We're observing what's in the world around us.
89
00:09:25,540 --> 00:09:29,229
We're making notes and we're starting to abstract, right?
90
00:09:29,230 --> 00:09:32,950
So this is what mathematicians do. They observe, they look for patterns.
91
00:09:32,950 --> 00:09:36,129
They say, Well, I'll make some notes, but I might come back that later.
92
00:09:36,130 --> 00:09:43,330
There seems to be some structure that might be interesting here. But the first thing I'd like to ask is, well, is the pitch in maths,
93
00:09:43,330 --> 00:09:49,569
in the fluent maths the same does having the same amount of symmetry does the fact I only had three symmetries,
94
00:09:49,570 --> 00:09:55,420
does that mean that if I find anything else with three symmetries and make notes, I'll have the same notes.
95
00:09:55,660 --> 00:10:00,910
Just the same amount of symmetry always mean you have the same type of symmetry.
96
00:10:01,510 --> 00:10:09,999
So I've got the question. So another way of saying that is, is number the correct counting, the correct way to try and talk about symmetry,
97
00:10:10,000 --> 00:10:15,250
is it enough to say, well, when we're discussing symmetry, we'll just count like we do with many things.
98
00:10:15,710 --> 00:10:21,180
What you need is there in more natural language that you should develop in order to properly discuss symmetries.
99
00:10:21,490 --> 00:10:26,969
Okay, so if everything only had three symmetries, it seems that well, it's probably they're probably all the same.
100
00:10:26,970 --> 00:10:30,970
But let's think of some other objects with three symmetries. Exactly.
101
00:10:30,970 --> 00:10:35,980
Three symmetries. Well, this is well, it's called a strange attractor. It arises in dynamical systems.
102
00:10:36,160 --> 00:10:41,920
It's a rather beautiful thing that, you know, that that that fluid will flow to if it's agitated in the right way.
103
00:10:42,520 --> 00:10:49,360
And that thing, again, has three symmetries, and it's obviously got more than a superficial resemblance to that flag up there.
104
00:10:49,540 --> 00:10:56,920
Okay. And that that's not because I went out looking for a representation of Manx ness in fluid dynamics.
105
00:10:57,250 --> 00:11:00,940
It's just because if you take something with three symmetries, that's what you'll find.
106
00:11:00,940 --> 00:11:09,400
You'll have the same sort of sense of affair, of rotation, of a three fold symmetry, just being some sort of rotation.
107
00:11:09,850 --> 00:11:14,590
That's another object with three symmetries. This time some Celtic not work.
108
00:11:15,040 --> 00:11:22,000
Again, it's not exactly three symmetries, but more than that it's got the same sort of feel that you just can rotate it right.
109
00:11:22,180 --> 00:11:25,180
And if you think of anything with three symmetry, that's what you're going to find, right?
110
00:11:25,420 --> 00:11:29,710
It's not just that has the same amount of symmetry, reason has the same type of symmetry.
111
00:11:30,970 --> 00:11:34,520
Well, I may or may not be interesting. Maybe we should look at a bigger example.
112
00:11:34,540 --> 00:11:40,420
How about six is a number? Okay. Well, there's lots of natural objects that have six symmetries.
113
00:11:40,810 --> 00:11:45,850
Fascinating fact is that if you take snowflakes, they tend to have this six fold symmetry.
114
00:11:46,720 --> 00:11:53,170
And so I call that a photo of a slop over that for reasons the experts know.
115
00:11:54,070 --> 00:11:58,750
Imagine that's a photograph of a snowflake. What can I do so that it's a real symmetry?
116
00:11:58,750 --> 00:12:03,060
So if you close your eyes, then open again. You won't notice I've done anything. Well, again, I can rotate it.
117
00:12:03,070 --> 00:12:07,080
Okay. I can rotate it once or twice or three clicks to four clicks.
118
00:12:07,080 --> 00:12:10,600
So five clicks, the six click will get me back to where I started from.
119
00:12:10,900 --> 00:12:15,879
Okay. And I might abstract that in all sorts of nice ways with the patterns again that
120
00:12:15,880 --> 00:12:20,950
come out of studying dynamical systems as a sort of a snowflake like object.
121
00:12:21,040 --> 00:12:24,070
And it's got the same sort of symmetry, right? Different object.
122
00:12:24,460 --> 00:12:29,620
But I can say exactly the same thing about the symmetries. Remember, I'm right in my notebook.
123
00:12:30,010 --> 00:12:37,749
So again, I notice I'm studying objects where you have a rotation that that somehow accounts for all the symmetry.
124
00:12:37,750 --> 00:12:41,530
That is to say, if I just keep doing this rotation, I'll get all the symmetries.
125
00:12:41,770 --> 00:12:46,720
And if I do it six times, I'll I'll get back to doing nothing if I'm happy.
126
00:12:47,110 --> 00:12:52,749
We just made similar notes as when we had three, but he's a different object now.
127
00:12:52,750 --> 00:12:57,069
That doesn't look the same. Okay, this is so we're not looking at nature.
128
00:12:57,070 --> 00:13:02,200
We're looking at snowflakes or or abstracted versions of snowflakes.
129
00:13:02,380 --> 00:13:05,710
Here's another object that goes nature, an ethane molecule.
130
00:13:06,760 --> 00:13:09,790
Okay, so now how many symmetries does that have?
131
00:13:10,870 --> 00:13:17,319
Well, you just stared. I don't will count again. Well, so these these big ones are carbon atoms.
132
00:13:17,320 --> 00:13:24,760
These these little ones are hydrogen atoms. So. So how can I turn it so that after I've turned it, it'll look the same?
133
00:13:24,760 --> 00:13:33,340
Well, I might if I decided I'll leave that carbon atom where it is, then I can obviously just rotate by a third of a rotation.
134
00:13:33,860 --> 00:13:38,890
Right. So I can rotate fixed that one, I can rotate those three around.
135
00:13:39,160 --> 00:13:46,240
So that would give me three possible symmetries, do nothing, rotate one third of a circle or take two thirds of a circle.
136
00:13:46,750 --> 00:13:53,260
And I call that t for twiddle or turn right and I could flip it right.
137
00:13:53,260 --> 00:13:56,230
I could just take hold of it and flip it around.
138
00:13:56,440 --> 00:14:04,750
So I interchange the two carbon atoms on that and that's essentially all I could do, except of course I could twiddle it and then flip it.
139
00:14:05,200 --> 00:14:08,529
And so you count how many things can go on here. Well, I can do nothing.
140
00:14:08,530 --> 00:14:15,250
I can always do nothing. Ha. Stay in bed in the morning or twiddle or twiddle twice.
141
00:14:15,620 --> 00:14:22,870
I can flip it like a toilet and then flip it or I could twiddle it twice and then flip it and you can convince yourself.
142
00:14:22,870 --> 00:14:27,550
But anything else you might try to do to that is on that list. So again, it has exactly six symmetries.
143
00:14:27,970 --> 00:14:32,890
Now, what am I going to write to my notebook this time? I'm going to say, well, I found two basic things you can do.
144
00:14:32,890 --> 00:14:37,870
You can twiddle or flip these two little three times that scratch back to do nothing.
145
00:14:38,140 --> 00:14:43,300
If you flip twice, that's obviously the same as doing nothing. That's what these notes mean.
146
00:14:44,190 --> 00:14:52,660
Then this time I'm going to note if I flip and then twiddle and then flip, it actually has the effect of twist twiddling in the opposite direction.
147
00:14:52,820 --> 00:14:56,200
There. Don't see that. Think about it. I should say.
148
00:14:56,200 --> 00:15:02,050
But I'm going to put these notes on on the Web so you can come back and look at these slides at your leisure.
149
00:15:02,290 --> 00:15:08,320
So. So there's another object six symmetries. But that doesn't look the same as the symmetries here.
150
00:15:09,280 --> 00:15:18,100
In particular, try and convince yourself that here you see this one basic operation, and I have to do it six times before I get back to doing nothing.
151
00:15:19,430 --> 00:15:26,960
Here. You convince yourself that anything you do, you either do it too, twice or three times before you get back to doing nothing.
152
00:15:27,050 --> 00:15:28,670
So that's something that's really different.
153
00:15:28,890 --> 00:15:36,740
Okay, so here we can see that these two objects have the same amount of symmetry, but they don't have the same type of symmetry.
154
00:15:36,740 --> 00:15:40,640
They don't have, in any sense the same symmetries.
155
00:15:40,940 --> 00:15:49,720
So so now we know that number counting is not enough of a language to help us describe the natural things we find in nature.
156
00:15:50,360 --> 00:15:54,990
Wanting to describe symmetry. We need some other structure. We need some better language.
157
00:15:56,120 --> 00:15:58,220
I'll just give you one example, which I'll do quickly,
158
00:16:00,290 --> 00:16:04,879
because I want to do this because let's think about the nature of symmetry, what I'm saying symmetry.
159
00:16:04,880 --> 00:16:07,400
And I hope you're not flinching at that. You know what I mean?
160
00:16:07,410 --> 00:16:12,380
I'm twiddling things around without breaking them so I can put them back before the museum curator comes.
161
00:16:13,940 --> 00:16:20,780
But according to the type of object you're interested in, what you mean by a symmetry will change, right?
162
00:16:20,930 --> 00:16:25,910
So if you've got a glass object, then you better not try to stretch it because it will break.
163
00:16:26,270 --> 00:16:32,299
So if symmetry is a rigid motion, right? If you have a rubber object, then you allow a symmetry to be.
164
00:16:32,300 --> 00:16:36,470
I might put my finger on it and stretch it a bit. Right, and then stretch it back.
165
00:16:36,590 --> 00:16:39,620
Right. That would be a perfectly good symmetry. I haven't broken anything.
166
00:16:39,620 --> 00:16:43,730
I could put it back. I didn't. I preserved all the nature of the thing.
167
00:16:44,360 --> 00:16:48,979
So what you think of as a symmetry depends entirely on what sort of thing you're doing.
168
00:16:48,980 --> 00:16:53,750
What sort of objects are you interested in? Okay, so in this case, I'll be very naive.
169
00:16:53,750 --> 00:16:58,430
I'll just have three balls. And now I'm interested in how can you rearrange three balls?
170
00:16:59,240 --> 00:17:03,200
Okay, so I don't. There's no rigidity now I can just move them around as I like.
171
00:17:03,770 --> 00:17:08,180
And so I claim once again, the answer is the six ways of doing this.
172
00:17:08,510 --> 00:17:11,569
You can do nothing or what I've called Alpha.
173
00:17:11,570 --> 00:17:18,620
You just take the first two and you interchange them or betas take the second two pair and you interchange them.
174
00:17:19,550 --> 00:17:25,700
And I claim that there are exactly six ways doing this. And if you work it out, if you alpha, you could do beta.
175
00:17:25,700 --> 00:17:35,180
If you do alpha then beta that has the effect of kind of cyclically permitting them or I claim the alpha beta alpha is the same as beta alpha beta.
176
00:17:35,440 --> 00:17:40,880
Okay. So once again, just check. It's lots of ways to convince yourself is exactly six ways of rearranging those balls.
177
00:17:41,600 --> 00:17:45,170
And, uh, and I'll make some notes on my notebook.
178
00:17:45,540 --> 00:17:53,150
I'm trying to develop a theory of symmetry in my notebook. And so this time I write down I've got two basic things I can do.
179
00:17:53,600 --> 00:17:57,889
They were flipped, but all I'm going to write down is the bare essentials because I don't have much space.
180
00:17:57,890 --> 00:18:03,440
My notebook, there's two basic operations, and I gave them Greek names and I observed again,
181
00:18:03,440 --> 00:18:07,760
I'll say, if you do eight alpha twice, beta twice, that's the same as doing nothing.
182
00:18:08,540 --> 00:18:12,410
And I observe this other little rule I noticed. Okay, because that seems sort of interesting.
183
00:18:12,680 --> 00:18:18,139
You write notes and things seem interesting. So now we've got three things that have six symmetry.
184
00:18:18,140 --> 00:18:21,200
So we've got that for which we had this description here.
185
00:18:22,190 --> 00:18:26,330
We got that's a more primitive one where we just rotate and we've got that object.
186
00:18:26,960 --> 00:18:30,030
Okay. Let's skip over that. Right.
187
00:18:30,600 --> 00:18:36,300
Okay. Now we reflect, right? We go and sit in the pub with our notebook and we think, what have we just been doing?
188
00:18:37,140 --> 00:18:45,090
We've been making notes about symmetries of things. But in that last one, we changed our mind about what we were, what we had symmetries of.
189
00:18:45,120 --> 00:18:48,660
So now we're going to do mathematics. We already are doing mathematics.
190
00:18:48,690 --> 00:18:51,990
Okay. Now, what's object? What? What's mathematics about?
191
00:18:52,110 --> 00:19:02,010
It's about studying objects called X. The only question you have to decide when you start doing mathematics is what is X?
192
00:19:02,490 --> 00:19:09,500
Right. You have to decide what interests you. And that that really is the start of a lot of mathematics.
193
00:19:09,510 --> 00:19:13,559
Right. So you observe patterns, you reflect on truth and beauty,
194
00:19:13,560 --> 00:19:20,630
and you think I care about these sort of problems and the objects that I want to think about have these properties.
195
00:19:20,640 --> 00:19:26,640
So I'm going to make a careful definition of those properties, and then I'm going to study all the objects to satisfy that.
196
00:19:27,120 --> 00:19:30,990
Okay. So maybe I'm interested in the geometry of rigid objects.
197
00:19:31,380 --> 00:19:34,650
Or maybe I'm interested in the geometry of rubber objects, how you can deform them.
198
00:19:35,100 --> 00:19:38,520
Maybe I'm just interested in counting things and moving them around.
199
00:19:39,000 --> 00:19:42,100
It's up to you. It depends entirely on what sort of problems you're trying to solve.
200
00:19:42,160 --> 00:19:48,450
The first thing you've got to do when you start being serious about mathematics is decide in order to do this type of mathematics,
201
00:19:48,450 --> 00:19:53,790
I'm going to study objects of this sort. You make a careful definition and that's your axes.
202
00:19:54,570 --> 00:19:58,320
And then no matter what you're doing, you sooner or later get interest in the symmetry.
203
00:19:58,320 --> 00:20:03,630
So that object. Okay. I'm interested in the properties of this object, the property that I use to define it.
204
00:20:04,200 --> 00:20:09,480
Now, what are all the things I can do to it that preserve all the properties I care about?
205
00:20:10,370 --> 00:20:14,190
Okay, so. So in the objects, we started with this on most primitive ones.
206
00:20:14,340 --> 00:20:19,230
You're interested in the ones you really can't see any difference. You preserve all the geometry when you change them.
207
00:20:19,560 --> 00:20:24,840
But no matter what sort of mathematics or science you're doing, you're really interested in the symmetries of your system.
208
00:20:25,780 --> 00:20:30,210
Okay. And so those symmetries. So, in other words, automotive isms.
209
00:20:30,540 --> 00:20:34,380
So I'm not going to use that word systematically, but it should be where it exists.
210
00:20:34,890 --> 00:20:38,760
All right. So you think about them. You got an object you care about now.
211
00:20:38,760 --> 00:20:42,090
You care about all the ways of preserving the essence of that object.
212
00:20:42,570 --> 00:20:49,200
So if you're just interested in sex buckets of balls or something with no extra structure, you're just interested in the ways of rearranging them.
213
00:20:49,770 --> 00:20:52,800
But if you're interested in glass objects, then you want to move them.
214
00:20:53,070 --> 00:20:58,440
So you preserve all distances. You don't break them. If you're interested in rubber objects, you might stretch them.
215
00:20:59,220 --> 00:21:04,380
And if you're interested in monolithic objects, you might require a finer structure, whatever.
216
00:21:04,950 --> 00:21:09,450
But just just just think I have objects, and there's ways of deforming them.
217
00:21:10,810 --> 00:21:14,220
Matt. This is the first point I really want to get to.
218
00:21:14,240 --> 00:21:21,080
So let all this wash over you, but then absorb this no matter what sort of mathematics you're interested in.
219
00:21:22,430 --> 00:21:25,550
The symmetries of your object. The old emotions. Your object.
220
00:21:25,790 --> 00:21:29,840
Always. Always. Always form something called a group.
221
00:21:30,690 --> 00:21:34,340
Okay. So this is a source of propaganda or.
222
00:21:34,520 --> 00:21:44,750
Or persuasion that groups are the appropriate mathematical language to study the symmetries of any object in any context.
223
00:21:45,200 --> 00:21:50,030
Now, for most, if a lot of primitive mathematics, the right concept is number you count,
224
00:21:50,600 --> 00:21:53,989
and then you develop number systems and you manipulate those numbers.
225
00:21:53,990 --> 00:22:00,610
Systems get tremendous mathematical power to solve questions that ask things like how much or how many?
226
00:22:00,770 --> 00:22:02,120
Or Is this greater than that?
227
00:22:03,050 --> 00:22:10,700
But when you're studying symmetry, automotive isms, symmetries of objects, the right question isn't it's a different nature.
228
00:22:10,700 --> 00:22:16,130
And numbers are not the right thing to talk about. You want to talk about these algebraic structures called groups?
229
00:22:16,490 --> 00:22:19,850
Now, this doesn't come out of thin air. It comes out of your experience.
230
00:22:20,210 --> 00:22:28,640
It comes out of your experience that we sort of were pretending with which we were running through examples of here you study objects,
231
00:22:28,820 --> 00:22:34,820
you make notes, and you start developing the sense of what the appropriate mathematical structure should be.
232
00:22:35,600 --> 00:22:43,730
So what is it, group? Well, you just try to make the best possible requirements that capture the essence that are common to all situations.
233
00:22:44,090 --> 00:22:47,720
But you care about symmetry. So what do you have to do?
234
00:22:47,750 --> 00:22:54,140
Well, as I said, you can always stay in bed. Every object of every sort has at least one symmetry.
235
00:22:54,260 --> 00:22:59,030
Just leave the damn thing alone. Just don't touch it. You certainly preserved all of its essence.
236
00:22:59,690 --> 00:23:08,530
So every in every setting, you should always have an operation, which is one for historical reasons, of just do nothing identity.
237
00:23:08,580 --> 00:23:11,720
Right? Just leave it alone. Stay in bed. Don't worry about it.
238
00:23:12,950 --> 00:23:17,210
And then the other thing you can do with what's a natural thing.
239
00:23:17,280 --> 00:23:23,000
If you if you to twiddle this in different ways, if you don't break it by doing that,
240
00:23:23,040 --> 00:23:27,170
you don't break it by doing that, then you can do that and that and you still haven't broken it.
241
00:23:27,500 --> 00:23:34,610
Okay. So you can compose when you've got two symmetries of an object, then you can do one followed by the other.
242
00:23:34,730 --> 00:23:36,560
And now you have a new symmetry of the object.
243
00:23:36,770 --> 00:23:44,180
So you have this notion of composing things, do A, then B, and if you didn't break it, then you can undo it.
244
00:23:44,210 --> 00:23:50,990
Right. Just run the movie backwards of you being caught on CCTV picking up that Ming vase in the museum.
245
00:23:51,300 --> 00:23:55,190
Right. As long as you can slowly and break it. As long as you've preserved all the essence.
246
00:23:55,400 --> 00:24:03,800
You can always undo what you did. So those are obvious inherent features of any system of symmetry, of any sort of object whatsoever.
247
00:24:04,100 --> 00:24:07,290
So we'll just make those our requirements, those and only those.
248
00:24:07,580 --> 00:24:14,270
Okay, that's a slight, right? But not really. And then what does it mean to say two things at the same?
249
00:24:15,300 --> 00:24:23,190
Well, if we were dealing in the realm of number, the appropriate notion of the same is having the same number of whatever as they are.
250
00:24:23,250 --> 00:24:31,680
Right. So we make the abstraction of two because we get fed up of saying that's the same that pair of apples and that pair of volunteers.
251
00:24:31,680 --> 00:24:34,950
That's the same number of apples as oranges. That pair of chestnut.
252
00:24:34,950 --> 00:24:41,550
Pair of apples. I can match them up so they have the same that there's the same amount of apples as there is of chazz.
253
00:24:42,270 --> 00:24:47,910
So we make the abstraction of a number. And then instead of having to compare it to something, we can then say there's two of them.
254
00:24:49,040 --> 00:24:55,290
This is basic mathematics. But you look at the common essence of a class of objects you're interested in, and you abstract it and you say,
255
00:24:55,410 --> 00:25:03,450
Now I have my ideal gold standard, one that I can leave in the Bank of England, and so that I can use this to to count pairs of things.
256
00:25:03,960 --> 00:25:09,000
And the same with groups instead of having to say, well, that's got the same type of symmetry as that,
257
00:25:09,240 --> 00:25:16,770
I want to make a list of groups and say, Well, whenever I find a new object, it ought to have symmetries from my list.
258
00:25:16,780 --> 00:25:21,330
Somehow I want to describe all possible symmetries of all possible objects.
259
00:25:21,930 --> 00:25:26,580
And when do you say two things are the same? Well, now it's not just a matter of matching them up.
260
00:25:26,910 --> 00:25:34,350
You should match them up in a way. So if I match A's with B's, A's being symmetries of one object to B's being symmetries of another,
261
00:25:34,860 --> 00:25:38,670
I should be able to match them up so that when I compose them,
262
00:25:38,850 --> 00:25:44,490
if I do a one of the A's and then another one on this side, it should be the same as that twin's on the other side.
263
00:25:45,240 --> 00:25:50,790
They won't have these. You have to match them up. Not just that they're counting the same, but they they respect the structure.
264
00:25:50,790 --> 00:25:56,520
We care about composing or inverting. Okay, that is the most technical bits.
265
00:25:56,530 --> 00:25:59,640
If you survive, that's okay. Just let it wash over you.
266
00:25:59,700 --> 00:26:03,629
If you didn't survive that, try to wake up now or go to sleep.
267
00:26:03,630 --> 00:26:07,710
As you wish. Right. But let's go back to our examples.
268
00:26:08,370 --> 00:26:11,430
So now we have a language we have the language of group theory.
269
00:26:11,770 --> 00:26:14,190
So now I'm going to just talk freely about groups. Okay.
270
00:26:14,190 --> 00:26:21,630
So member groups, a group is a system of symmetry that describes the symmetries of some object.
271
00:26:21,960 --> 00:26:26,580
But I've abstracted so just like I abstract from counting to having a notion of number.
272
00:26:27,150 --> 00:26:34,170
And I can then talk about numbers. Now, instead of having to talk about the symmetries of a specific object, I can just talk about groups.
273
00:26:34,800 --> 00:26:38,660
Okay. Oops. I didn't wanna do that.
274
00:26:38,990 --> 00:26:44,210
Let's go back to our three basic objects. We had this bucket of three balls with those symmetries.
275
00:26:44,570 --> 00:26:47,990
We had our cyclic thing. Except I traded in four different cyclic things.
276
00:26:48,530 --> 00:26:55,700
And when I made those notes and I had my ethane molecule where I had those notes back groups.
277
00:26:57,370 --> 00:27:02,620
Now I have three different descriptions of groups. Remember here I said there's two basic things you can do.
278
00:27:02,920 --> 00:27:07,829
Here are some observations about how these things behave. Here.
279
00:27:07,830 --> 00:27:11,010
I said, Well, there's one basic operation that gives you the whole shebang.
280
00:27:11,280 --> 00:27:14,790
And the only thing I know is if you do it six times, it's the same as doing nothing.
281
00:27:15,120 --> 00:27:20,110
And here I made some different notes. Now all those groups are the same.
282
00:27:20,710 --> 00:27:24,280
Well, we sort of argued earlier that that one was different to that one.
283
00:27:25,120 --> 00:27:29,560
Okay. But in fact, those are the two. That one and that one are the same.
284
00:27:30,040 --> 00:27:36,819
Okay. So that's just a little exercise. We've never seen it. Think about how to match up the symmetries of that object with the symmetries of
285
00:27:36,820 --> 00:27:40,840
rearranging three balls in such a way that you expect all the compositions and things.
286
00:27:42,470 --> 00:27:43,510
Okay, that's nice.
287
00:27:43,540 --> 00:27:51,490
On one hand, we sort of getting somewhere and be able to describe groups, but we're probably starting to be a little bit worried because.
288
00:27:52,700 --> 00:27:54,860
It's a moment at which you should not feel alone.
289
00:27:55,820 --> 00:28:02,390
If somebody just walked into the room and said, I found this lovely object and here, here are the notes I made about cemeteries.
290
00:28:03,170 --> 00:28:06,499
And somebody else came in and said, Well, oh, I found this other object.
291
00:28:06,500 --> 00:28:09,710
I don't think it's like yours. But here's the notes I made about its symmetries.
292
00:28:10,700 --> 00:28:17,390
And he said, Well, it's your system of symmetry, sort of the same as mine, although we talk about the same group in a different language.
293
00:28:18,340 --> 00:28:26,230
Well, the answer is yes, but it's not obviously yes right now that's going to worry us because we got so used to number two.
294
00:28:26,290 --> 00:28:31,060
So when I say two and you say two, we have no doubt that we talk about the same thing.
295
00:28:31,330 --> 00:28:34,750
But now when we talk about groups, we've described the same group,
296
00:28:34,750 --> 00:28:39,400
quite an easy group, just something with six operations in it in two different ways.
297
00:28:40,210 --> 00:28:45,250
And that's making us nervous to think that actually recognising what you've got might be a bit tricky.
298
00:28:46,240 --> 00:28:53,559
Okay. Hold that thought. Okay, so we've got a language for describing symmetry, but we're a bit worried that when we have different descriptions,
299
00:28:53,560 --> 00:28:56,860
it's not very easy to see if we've really got the same thing in disguise.
300
00:28:58,860 --> 00:29:01,979
Okay. Well, we've counted. We've done some finite things.
301
00:29:01,980 --> 00:29:07,530
Let's do some infinite things. We feeling strong enough for infinite things. Some of you look strong enough for infinite things.
302
00:29:07,530 --> 00:29:11,220
Right? So imagine this, the paving going on forever.
303
00:29:11,580 --> 00:29:15,149
Okay, here's a nice, simple, infinite system of symmetry.
304
00:29:15,150 --> 00:29:19,320
An infinite group. You look at that paving, stare at it for a while.
305
00:29:19,320 --> 00:29:21,050
You realise there's two nice things you can do.
306
00:29:21,060 --> 00:29:27,480
You can shift this, this paving stone I'm pointing out along to that lawn and the whole system will repeat.
307
00:29:28,330 --> 00:29:34,590
Okay, so that's a paving. That will be a symmetry of that paving. Imagine this paving goes on forever.
308
00:29:36,180 --> 00:29:41,100
It's a mathematics institute, not an engineering one. So imagine this paving goes on forever.
309
00:29:41,100 --> 00:29:45,600
That's obviously a repeating pattern shifted along. Or you can shift up.
310
00:29:45,600 --> 00:29:48,870
You can shift that one up to that. And it's doubly periodic.
311
00:29:48,870 --> 00:29:52,440
It's periodic in that direction and that direction. Everybody happy?
312
00:29:52,880 --> 00:29:59,340
Those two basic operations, I'll call them A and B, and the only note I'm going to make is, well,
313
00:29:59,340 --> 00:30:05,070
look, if I shift along and then up, that has the same effect as if I shift up and then across.
314
00:30:05,770 --> 00:30:08,270
So that's the only note I'll make as I've described myself.
315
00:30:08,280 --> 00:30:15,900
Another group of note to this two basic operations that seem to give me all the symmetries if I repeat them and I only notice one rule,
316
00:30:17,400 --> 00:30:23,340
it's a rather nicer pattern. This is a Charles Rennie Mackintosh wallpaper.
317
00:30:23,340 --> 00:30:26,710
I think now it's different.
318
00:30:26,730 --> 00:30:30,270
The previous object, right. But clearly got the same symmetries.
319
00:30:30,720 --> 00:30:38,550
All the symmetries of this again, what can I do? Well, I can either shove it along one column to the next, I can shove it up by two rows.
320
00:30:39,930 --> 00:30:44,880
And again, I'll call those symmetries and B and the only thing to really note is that doesn't matter what order I do them in.
321
00:30:46,050 --> 00:30:53,040
Now, that's clearly the same group. Okay. And it doesn't come as any surprise to you say that that has the same symmetries,
322
00:30:53,040 --> 00:30:57,750
is that they're different objects, of course, but it's intuitively clear they have the same symmetries.
323
00:30:57,960 --> 00:31:01,350
Everybody happy, right? How about that?
324
00:31:02,340 --> 00:31:05,700
Okay. Well, that's got a few extra symmetries, right?
325
00:31:05,700 --> 00:31:11,459
So so that is rather like the previous one. It's just a tiling of some plainer surface.
326
00:31:11,460 --> 00:31:14,880
Again, you can again shove it along by one column.
327
00:31:15,330 --> 00:31:19,380
That's a perfectly good symmetry. Observe all the shapes and all the colouring.
328
00:31:20,280 --> 00:31:22,740
So you have to show off that little. You have to shove that.
329
00:31:23,280 --> 00:31:29,460
So you have to shove that point over to that, get a symmetry and you shove it along does that, or you can shove it up.
330
00:31:29,610 --> 00:31:32,670
It has those symmetries again, but now it has a bit more symmetry.
331
00:31:33,540 --> 00:31:38,219
So how are we going to describe it this time? Well, I want you to focus on some particular point.
332
00:31:38,220 --> 00:31:41,760
Let's take that one. Let's take that little purple triangle there.
333
00:31:42,990 --> 00:31:52,500
Now, I have to get this order right here. If I if I focus at that corner of the triangle, and then if I turn it through 180 degrees.
334
00:31:53,740 --> 00:32:00,340
Right. If I stick my finger in the pot and then turn it by 108 degrees, that will be a nice symmetry of the pattern.
335
00:32:00,820 --> 00:32:06,340
I'll call that a and if I do it twice, I'll get back to I started from I just rotate by 180 degrees.
336
00:32:06,880 --> 00:32:11,200
If I go to this corner, then I have to rotate by a third of a rotation.
337
00:32:11,530 --> 00:32:17,770
I'll call that B, right. So that operation stick your finger there and rotate by a third of a circle.
338
00:32:17,920 --> 00:32:20,560
So if you do that three times, you obviously get back to I started from.
339
00:32:21,190 --> 00:32:27,760
And if you go to the final corner of the triangle, that one there, then you have to rotate by you can rotate by a sixth of a circle.
340
00:32:27,970 --> 00:32:31,230
And if you rotate six times, you can actually start from there.
341
00:32:31,630 --> 00:32:36,970
So I've spotted three extra symmetries that aren't to do with pushing it around, but now it's doing rotating.
342
00:32:37,420 --> 00:32:38,979
And I'll make some little notes myself.
343
00:32:38,980 --> 00:32:47,410
Look, I found these three cool things, like B and C, if I do A twice or B three times, A 66 times the same as doing nothing.
344
00:32:47,980 --> 00:32:51,730
And I make one other observation. This is one for you to check at home.
345
00:32:52,060 --> 00:32:57,760
If you do A, then B, then C you'll get back to I started from everything will be moved back to where it started from.
346
00:32:59,390 --> 00:33:02,660
Now I've made some observations about the symmetries of this,
347
00:33:02,660 --> 00:33:06,620
and I've made notes in my famous notebook that I have to take to the pub and reflect on later.
348
00:33:07,760 --> 00:33:11,650
Now, have I captured all of the cemeteries of this in particular?
349
00:33:11,660 --> 00:33:15,380
What happened to that? Shoving it along a bit and shoving it up a bit?
350
00:33:16,970 --> 00:33:20,090
When I say I've made these notes and I've got it right.
351
00:33:20,480 --> 00:33:24,950
Well, I mean, two things I'm asserting, first of all,
352
00:33:25,730 --> 00:33:34,250
that every possible symmetry of that picture is obtained by doing these basic operations A, B and C repeatedly.
353
00:33:34,460 --> 00:33:43,220
Maybe I'll do A once and then B twice, then C twice, then I'll undo A, then I'll do B twice, then I'll do C three times.
354
00:33:43,550 --> 00:33:52,520
So you just make combinations, these basic operations exercise for all those bright eyed young students over there and prove
355
00:33:52,610 --> 00:33:58,010
that every way of moving this pattern to itself can be obtained by repeating those operations.
356
00:33:58,160 --> 00:33:59,330
That's the first assertion.
357
00:34:00,440 --> 00:34:10,729
Second assertion, any rule that you tell me, connecting some combination of B and C, so you say, look, man, I've I've been working on this all night.
358
00:34:10,730 --> 00:34:14,030
I can see that this massively long combination of AP and CS.
359
00:34:14,240 --> 00:34:17,270
Well, in an unexpected way. Get me back to where I started from.
360
00:34:17,960 --> 00:34:22,790
I claim. Oh, I knew that. I would say that, wouldn't I?
361
00:34:22,970 --> 00:34:30,290
So I claim that anything you can tell me about combinations of AP and C follows from these simple rules here.
362
00:34:30,920 --> 00:34:36,590
Okay. Now, I'm not going to delve into what I mean, but I honestly mean that literally anything you can prove,
363
00:34:37,070 --> 00:34:40,280
any combination of moves that you can prove are related to each other,
364
00:34:40,280 --> 00:34:45,860
any equation relating A, B and C can be deduced from these four simple equations here.
365
00:34:47,820 --> 00:34:51,690
So I have nice notes explaining exactly the symmetries of that object.
366
00:34:51,870 --> 00:34:58,770
Well, that's good, right? So what I'm getting to is my notes I've been making all along, my famous notebook.
367
00:34:58,950 --> 00:35:02,190
I've been jotting down the symmetries of every object I come across.
368
00:35:02,460 --> 00:35:09,150
And I have developed a sort of calculus, a sort of notation that can let me talk about all the symmetry groups I ever come across.
369
00:35:09,690 --> 00:35:15,059
And what I'm going to do is I'm going to make notes. I have some basic transformations, some basic operations.
370
00:35:15,060 --> 00:35:22,390
I do my thing. When I have my fingers, I'll make some notes, some equations about some rules that are satisfied.
371
00:35:22,410 --> 00:35:25,680
Like here, I've noticed that these little rules are satisfied.
372
00:35:26,220 --> 00:35:33,600
Excuse me? Are satisfied. And I'll start making notes when I'm convinced that everything,
373
00:35:33,850 --> 00:35:41,850
every possible equation relating my basic operations can be deduced in a purely formal way from the ones I've written down.
374
00:35:42,270 --> 00:35:47,430
Okay. So for example, let me give a simple example. If you tell me, look, I'm a clever boy.
375
00:35:47,490 --> 00:35:52,740
I've noticed that if you do a four times, it's the same as doing it, doing nothing.
376
00:35:53,250 --> 00:35:56,579
Well, I'll say of course it is, because if you do a twice, it's already nothing.
377
00:35:56,580 --> 00:36:04,290
And if you do nothing twice, it's the same as doing nothing. And so, I mean, that is that sort of obvious deduction from the rules.
378
00:36:04,290 --> 00:36:09,810
Anything you can say about the linking A, B and C can be deduced from these simple rules.
379
00:36:10,410 --> 00:36:14,460
And so I'm interested in making accurate notes in that sense.
380
00:36:15,030 --> 00:36:18,420
You have to have enough basic operations to generate everything,
381
00:36:19,020 --> 00:36:25,890
and you have to have made enough notes to have information that allowed you to use any equation that holds between these generators.
382
00:36:27,180 --> 00:36:35,880
Now. So that means we can start just writing down systems of symmetry.
383
00:36:36,360 --> 00:36:39,440
So knowing the theory is a very powerful thing, right?
384
00:36:39,450 --> 00:36:42,509
Once you've decided I know what I mean by numbers,
385
00:36:42,510 --> 00:36:50,250
you start manipulating them and you get you find fine structure that will tell you about real world objects.
386
00:36:51,150 --> 00:36:57,330
I now have a we now have a way of just describing systems of symmetry.
387
00:36:57,820 --> 00:37:06,660
Right. So you might before you ever meet something with 1,000,000,517 objects in it, you sort of have a conception of what that number is.
388
00:37:06,810 --> 00:37:11,490
Right. And you believe there are things that that have exactly that many members.
389
00:37:12,330 --> 00:37:18,240
So now how about if I just write down a system of symmetry and say, well,
390
00:37:18,420 --> 00:37:22,590
that must be the symmetries of something because it's a perfectly good system of symmetry.
391
00:37:23,130 --> 00:37:27,730
So, for example, his four things I might write down, I'll say I'll,
392
00:37:28,020 --> 00:37:35,940
I'll look for objects that have two basic symmetries, A and B, and it doesn't matter what order you do them in.
393
00:37:35,970 --> 00:37:40,379
Be followed by A is the same as a fall by B. Well, we saw objects like that, right?
394
00:37:40,380 --> 00:37:42,960
Our wallpaper and our paving were objects like that.
395
00:37:43,530 --> 00:37:49,560
But now I'll write down some more interesting rules, and then I'll go and seek objects that have those symmetries.
396
00:37:50,340 --> 00:37:54,360
And what I'm hoping is, well, I love geometry,
397
00:37:54,360 --> 00:38:01,650
so if I write down some system of symmetry and go hunting in some intelligent way, there's something that has those symmetries.
398
00:38:01,860 --> 00:38:10,350
Maybe I'll discover new geometries, objects that really have interest, partly because they have these interesting systems of symmetries.
399
00:38:11,220 --> 00:38:18,660
So how about this one? So instead of saying it doesn't matter what order do, and B, and suppose I have two basic objects.
400
00:38:19,110 --> 00:38:26,570
And I'll say if you do B that A, it's not the same as doing A than B, it's actually the same as doing A twice than B.
401
00:38:27,110 --> 00:38:31,410
I had to write this down so I didn't get it wrong in the wrong order. So.
402
00:38:31,530 --> 00:38:37,230
So can you think of anything like that? Okay. Well, here's his here's a couple of simple things you can do.
403
00:38:38,790 --> 00:38:47,120
Take a line. Okay. Which one's which yet? And say A is the one that translates this line to the right.
404
00:38:47,130 --> 00:38:54,870
So it sends X to X plus one. Okay. So any point on this line, it moves it along to X plus one is move the line rigidly.
405
00:38:55,770 --> 00:38:58,950
Okay. Now, let's mark something called zero on this line.
406
00:38:59,130 --> 00:39:06,780
And let's say B is the thing which contracts the line by a factor of two towards the origin.
407
00:39:07,310 --> 00:39:10,700
Okay, so imagine this line is made of rubber. You can shrink it by a factor of two.
408
00:39:10,710 --> 00:39:18,990
I can move it along by one. I leave it up to you to check that if you do B, then a a has the same effect as doing eight twice.
409
00:39:18,990 --> 00:39:24,170
And then b. So found an object that has those symmetries.
410
00:39:24,340 --> 00:39:28,090
He has two symmetries of a rubber line that satisfies the one we want.
411
00:39:28,300 --> 00:39:34,480
We've got that group that was just given to us by some some some abstract dry algebra.
412
00:39:34,510 --> 00:39:37,900
And now we've conjured up an object that has the right symmetries.
413
00:39:38,980 --> 00:39:42,730
Another thing you can do, by the way, for those of you who know what matrices are,
414
00:39:44,140 --> 00:39:52,180
if you take B to be the matrix like that, you take A to B, the matrix like that.
415
00:39:53,530 --> 00:40:03,730
Okay. So these are ways of moving the the the to the plane to itself, that all lines go to lines and you can find those matrices satisfy these rules.
416
00:40:04,270 --> 00:40:07,180
So that's a different object that has the same symmetries, right?
417
00:40:07,420 --> 00:40:12,219
You can take a rope a line or you can take a plane and transform it in a way that always sends lines.
418
00:40:12,220 --> 00:40:17,110
Two lines is two ways in which you can realise this abstract system of symmetry.
419
00:40:19,170 --> 00:40:22,200
Right. This is a moment to pay close attention.
420
00:40:22,710 --> 00:40:27,990
And I hope you're not colour-blind. The red group. Sorry, the red group.
421
00:40:27,990 --> 00:40:31,800
But don't worry. It's called something different. G three is red. G four is blue.
422
00:40:32,550 --> 00:40:38,730
Now they're modelled on this one, right? So now got three basic operations and they sort of follow the same rules.
423
00:40:39,510 --> 00:40:45,330
Here. I've got four operations. They follow the same rules. Let's hunt for objects that have these symmetries.
424
00:40:46,770 --> 00:40:50,490
Okay, so maybe it's like this. Maybe we'll be able to do something just for the line.
425
00:40:50,880 --> 00:40:56,490
Well, maybe we'll be able to write down some matrices and some dimension, maybe some bigger matrices because the bigger groups.
426
00:40:56,790 --> 00:41:04,010
And we'll look for ones to satisfy those rules. Hands up if you think this will be successful.
427
00:41:04,960 --> 00:41:08,500
My. Might be right.
428
00:41:08,770 --> 00:41:13,630
It might be. Might be. It won't. So instead.
429
00:41:14,810 --> 00:41:19,670
Maybe I'll just look for some finite object and look for some symmetries of this finite object,
430
00:41:19,670 --> 00:41:25,040
like maybe some nice big a molecule and ethane, and see if it has three symmetry satisfying these rules.
431
00:41:27,280 --> 00:41:31,100
Again, you'll fail. Now. Why will you fail?
432
00:41:31,130 --> 00:41:32,510
How do I know you will fail?
433
00:41:32,750 --> 00:41:44,570
Because it's I know the following non-obvious fact, but neither of these systems of symmetry arises as the symmetries of any finite object.
434
00:41:45,440 --> 00:41:49,429
Nor can it be realised as a group of matrices.
435
00:41:49,430 --> 00:41:54,410
You can't get it to act on any linear space and any way that preserves lines in any geometry.
436
00:41:56,270 --> 00:42:06,020
Now, in one case, there's a very good reason for that. One of those groups is the trivial group, which is to say from the rules I've written down.
437
00:42:06,470 --> 00:42:11,540
So maybe it's that one. Maybe it's that one. Let me just pick on this one arbitrarily.
438
00:42:12,380 --> 00:42:18,110
I if it if this group is trivial, I would say you've written down a system that's just stupid.
439
00:42:18,260 --> 00:42:21,350
It cannot be the systems, the symmetries of any object.
440
00:42:21,740 --> 00:42:22,760
How do I know that?
441
00:42:23,000 --> 00:42:32,420
Because by a long and complicated calculation, I would figure out that these rules imply that alpha, beta, gamma and Delta are all the identity.
442
00:42:32,420 --> 00:42:35,520
They all do nothing. Now.
443
00:42:35,520 --> 00:42:39,810
Maybe I'm lying to you. Maybe that's not true of that one. Maybe that's true of this one.
444
00:42:41,430 --> 00:42:46,170
Well, actually, only one of those groups is the trivial system of symmetry.
445
00:42:46,650 --> 00:42:52,290
The other one is an infinite group. That is to say, it's the symmetries of an infinite object.
446
00:42:52,500 --> 00:42:55,410
But it's not the symmetries of any finite object.
447
00:42:56,140 --> 00:43:02,290
And has no finite dimensional manifestations of its manifestations on infinite objects and infinite dimensions.
448
00:43:03,670 --> 00:43:14,080
Okay, let's have a vote. Who thinks G3 is trivial but G4 is infinite.
449
00:43:15,840 --> 00:43:19,990
You'll have to vote one way or the other. That's always interesting. Okay. Thank you.
450
00:43:20,860 --> 00:43:24,460
I forgot what I said. I said that was that was trivial.
451
00:43:24,490 --> 00:43:28,420
I said, okay, who thinks G4 is trivial but G3 is in front.
452
00:43:30,110 --> 00:43:36,139
It's always the same, it's always more or less 5050 it's actually G three that's trivial.
453
00:43:36,140 --> 00:43:39,560
And G four is an infinite group that has no finite manifestations.
454
00:43:40,190 --> 00:43:43,880
If you've got it wrong, don't worry. This is not obvious.
455
00:43:43,950 --> 00:43:53,390
Right now this is a manifestation. So that's not a very those aren't very big examples and it's already extremely hard to tell which is which.
456
00:43:54,050 --> 00:43:57,860
Now, what we're looking at here is a problem, a decision problem.
457
00:43:58,370 --> 00:44:01,729
Now, it's not undecidable that I knew the answer, for example.
458
00:44:01,730 --> 00:44:09,469
So can't be that undecidable, but it's a manifestation of the fact that problems about even fairly small group
459
00:44:09,470 --> 00:44:14,270
presentations can be intensely difficult and in fact generally impossible.
460
00:44:14,270 --> 00:44:17,750
And I want to explain what I mean by that impossibility, that underside ability.
461
00:44:18,680 --> 00:44:26,420
This is a picture of Max Day, the beginning of the 20th century, and he was the first one in this famous paper who realised there's a problem here.
462
00:44:27,140 --> 00:44:32,780
And what he saw, what he said was, was all very well writing down systems of symmetry and then saying,
463
00:44:32,780 --> 00:44:38,120
now I know everything about it because I've made clever notes before I went to the pulp and I have a complete description of this group,
464
00:44:38,660 --> 00:44:42,470
but it's very hard to see what information is written down.
465
00:44:42,680 --> 00:44:48,620
It's very hard to extract real information from those notes, even if in theory you've got complete information.
466
00:44:49,430 --> 00:44:55,370
So, for example, so what he asked is, is it actually impossible and do that actually exist?
467
00:44:55,370 --> 00:44:59,630
Algorithms, processes that you could always be confident would work,
468
00:45:00,500 --> 00:45:07,100
always be sure would work that would allow you to answer basic questions, such as is the group I gave you trivial?
469
00:45:07,100 --> 00:45:12,319
Is there really nothing it could possibly be the symmetries of or whether certain
470
00:45:12,320 --> 00:45:15,920
equations I give you can really be deduced from the relations that you think it can.
471
00:45:17,280 --> 00:45:23,249
Okay. So those are those illustrate the fact that it's intensely difficult to take the
472
00:45:23,250 --> 00:45:26,700
information out of out of systems of symmetry that are written down like that.
473
00:45:27,540 --> 00:45:30,450
But I want to pick up on why Dane was interested in this.
474
00:45:31,380 --> 00:45:38,670
Dane wasn't just interested in the way I described as just thinking of symmetries of things and coming to it from a algebraic point of view.
475
00:45:39,030 --> 00:45:45,810
What Dane was trying to do was understand all models of possible three dimensional space, so called three dimensional manifolds.
476
00:45:46,200 --> 00:45:51,060
And he was completely particularly interested in knots. So look at this comment.
477
00:45:51,540 --> 00:45:55,920
One is led to such problems by necessity, and it really is by necessity.
478
00:45:56,220 --> 00:46:01,290
When working in geometry, in topology that knotted space curves in order to be completely understood.
479
00:46:01,560 --> 00:46:04,830
The man solution of these of these algebraic problems.
480
00:46:05,640 --> 00:46:10,320
So let me try. So I'm going to explain why that is, but I'm going to try and illustrate it so that let it wash over you.
481
00:46:10,710 --> 00:46:15,120
The following slide is the right impression. Oh, not that slide. There are some knots.
482
00:46:16,320 --> 00:46:20,040
There is some knot. Now that that is not table meaning.
483
00:46:20,220 --> 00:46:24,000
All of those knots are different. Now, those are the simplest knots.
484
00:46:24,000 --> 00:46:29,280
How many there are that is overall is six, eight, 15, is it?
485
00:46:30,450 --> 00:46:33,960
Those are the 15 simplest knots in some sense. Okay.
486
00:46:35,900 --> 00:46:40,220
Now. How do you know that? Different. That is to say, if I take two of those.
487
00:46:40,340 --> 00:46:44,900
How do I know I can't take a piece of string depicted in one of them and manipulate it to be the other?
488
00:46:46,300 --> 00:46:49,600
How do you know that's an honest list with no redundancy on it?
489
00:46:50,770 --> 00:46:53,979
That's exactly the sort of thing Dane was interested in. Okay.
490
00:46:53,980 --> 00:46:58,900
How can you tell that not that appear to be different or space appear to be different, really are different.
491
00:46:59,710 --> 00:47:04,690
Now, people are very good at this. In the 19th century, though, he here's a 19th century not table.
492
00:47:04,990 --> 00:47:10,990
Actually, I think this is exactly the one that was on the murderer's wall in the last episode of Lewis.
493
00:47:13,230 --> 00:47:16,629
Yeah. So he's not.
494
00:47:16,630 --> 00:47:22,570
So how are you going to tell those the different. Well, you think maybe that's not so hard, but let me show you some examples.
495
00:47:23,860 --> 00:47:29,980
You have to stare at that picture for quite a while to see that if you give it a good tug, you cannot, not it.
496
00:47:30,880 --> 00:47:42,280
That really is the are not disguised. This one is the example of Wolfgang Harken that is also the UN, not if you pull out around enough,
497
00:47:42,400 --> 00:47:45,910
you'll find that you can deform it without passing it through itself to a circle.
498
00:47:46,270 --> 00:47:51,579
Now, that's not obvious, right? So Dane, being a clever guy, I thought, I know what I'll do.
499
00:47:51,580 --> 00:47:56,590
I'll translate this into an algebraic problem. So we associated a presentation of a group.
500
00:47:56,800 --> 00:48:00,760
He associated a group with the sort of description I've been given to each not.
501
00:48:01,330 --> 00:48:04,930
And he then thought he proved you actually made a mistake, that somebody fixed it later.
502
00:48:05,170 --> 00:48:12,040
He thought it proved that if you can say something sensible about the knot about the group, you'll know exactly which knot you've got.
503
00:48:12,760 --> 00:48:18,790
And this is so you see lots of theorems in mathematics to say I have reduced this difficult problem to this much easier one.
504
00:48:19,870 --> 00:48:25,929
But one should always beware of such statements. There's some sort of the some sort of principle of preservation, of hardness.
505
00:48:25,930 --> 00:48:27,160
And this is a good example.
506
00:48:27,600 --> 00:48:32,860
After thinking about it for a while and realise that the problem it translates into is harder than the one he started with.
507
00:48:33,940 --> 00:48:41,140
So it's hard to tell that that is the UN, not that it's really a circle and it's extremely hard to see that its group is the group of the unknown.
508
00:48:42,040 --> 00:48:46,780
Oh dear. This is all getting a bit bleak. So here's the underside of ability.
509
00:48:46,970 --> 00:48:50,799
So. So these are these aren't utterly esoteric problems, right?
510
00:48:50,800 --> 00:49:00,280
You want to know what a nut is? You want to know what a group is. And it was unknown from the time of day until the very end of the 20th century,
511
00:49:00,730 --> 00:49:05,470
whether it's actually possible in an algorithmic way to declassify or not.
512
00:49:05,500 --> 00:49:13,010
So you have these not tables that go up to about 14 crossings, but it is actually unknown whether it's an algorithmically solvable problem.
513
00:49:13,010 --> 00:49:16,990
If somebody gives you two pictures of or not, can you decide to the same knot or not?
514
00:49:18,560 --> 00:49:22,610
Not the UN not, not. Not yet. How do you decide if to not two different.
515
00:49:23,110 --> 00:49:30,070
Okay. It turns out and I'll try and explain in the next few minutes that that problem is solved.
516
00:49:30,080 --> 00:49:36,500
But it uses a huge amount of mathematics throughout the 20th century to answer that simple question.
517
00:49:36,500 --> 00:49:39,710
Can you distinguish between knots in a systematic way?
518
00:49:40,430 --> 00:49:46,160
What came earlier, what came in mid-century was the realisation that the algebraic version of that problem,
519
00:49:46,490 --> 00:49:51,590
these basic problems that Dave ran into of Can you tell what group you've got when you write down
520
00:49:51,590 --> 00:49:57,020
a system like can you tell if that's the trivial group that is algorithmically undecidable?
521
00:49:58,170 --> 00:50:05,030
Okay, now what? That me and I'm not not that we haven't yet invented an algorithm or the techniques we have so far insufficient.
522
00:50:05,450 --> 00:50:12,440
But one can prove that the can never exist a systematic way of deciding if a group is the trivial group or not.
523
00:50:13,100 --> 00:50:19,220
Okay. You really, really. You might at this point think, well, I'm smarter than some dumb machine.
524
00:50:19,230 --> 00:50:27,030
I could decide. No, no, you can't. Right. So anything that you could reasonably do could be done by a computer.
525
00:50:27,180 --> 00:50:30,800
And so there's a real theorem here that says there is no algorithm that,
526
00:50:30,840 --> 00:50:34,590
given an arbitrary, finite group presentation, will certainly give the correct answer.
527
00:50:34,770 --> 00:50:42,820
Yes, it's trivial. Oh, no, it's not. Now, that doesn't mean if you're given a particular instance of a problem, you can't solve it.
528
00:50:43,000 --> 00:50:49,980
What it means is, before knowing which which group I'm going to give you, you can't be sure that if you give me that, you can't think.
529
00:50:49,990 --> 00:50:54,280
If he gives me that, I'm going to do this, this and this, and I'm sure it'll work. You can never be sure.
530
00:50:54,560 --> 00:51:00,520
Okay, so this is undesired ability. Now this isn't way out.
531
00:51:01,060 --> 00:51:05,020
So I want to convince you this isn't this way out. An idea as you might think. So you might think.
532
00:51:05,020 --> 00:51:08,600
Well, decide if an arbitrary thing is trivial. That's undecidable.
533
00:51:08,710 --> 00:51:12,440
Okay. It took until 2050.
534
00:51:12,460 --> 00:51:15,340
Well, it took a few years ago, but something just appeared in the literature.
535
00:51:15,520 --> 00:51:22,960
So Welter and I proved if I if you like those groups, G three and G four, I had if you give me a group and ask,
536
00:51:23,170 --> 00:51:29,230
can this be realised of the symmetries of some finite object, we now know that's an undecidable problem.
537
00:51:30,040 --> 00:51:36,040
We cannot tell, given a system of symmetry, whether as a finite object with symmetries of the prescribed type.
538
00:51:36,640 --> 00:51:39,310
And here's an even perhaps more concrete term anyway.
539
00:51:39,320 --> 00:51:50,020
So Andrew Wiles up there, of course, is famous for telling us that if you pick an integer that's bigger than two and you try to solve the equation,
540
00:51:50,020 --> 00:51:54,610
let's pick L1, all the integers X, Y and z non-zero integers.
541
00:51:54,820 --> 00:51:57,400
So that's fine. That equation. And famously the answer's no.
542
00:51:58,480 --> 00:52:06,549
But one might hope that with advances in technology, eventually, if I give you any polynomial, it's called diophantine equations,
543
00:52:06,550 --> 00:52:13,240
a polynomial equation with integer coefficients and ask you Does that have any nontrivial solutions or not?
544
00:52:13,930 --> 00:52:19,660
You might hope that eventually the theory would develop far enough that no matter what equation I give you, you'll be able to tell me yes or no.
545
00:52:20,260 --> 00:52:27,640
Right? That's never going to happen. But not you say, which proved in 1970 that that sort of diophantine problem,
546
00:52:27,820 --> 00:52:35,260
that very concrete problem of given a polynomial and several variables with integer coefficients, does it have a solution in integers?
547
00:52:35,770 --> 00:52:41,530
That's an undecidable problem. So as Andrew proved, for certain equations you can decide, right?
548
00:52:41,770 --> 00:52:46,000
But if you don't know what equation I'm going to give you, there's nothing that you can hope.
549
00:52:46,240 --> 00:52:50,950
You can be confident will work in every case. Okay.
550
00:52:50,960 --> 00:52:56,000
So we've talked about symmetry and we've talked about underside ability.
551
00:52:56,520 --> 00:52:58,819
Okay. We haven't talked so much about geometry.
552
00:52:58,820 --> 00:53:05,630
And I realised that there are halfway through my slides and I have 5 minutes to go and I'm not going to rush, but I do want to talk.
553
00:53:05,870 --> 00:53:10,600
So we've had a lot. So. We need action.
554
00:53:11,080 --> 00:53:14,470
It's no use just sitting there. We need action.
555
00:53:14,800 --> 00:53:19,190
It's no use. Just sitting there staring at these algebraic descriptions of systems of symmetry.
556
00:53:19,660 --> 00:53:23,350
You've got. You've got to get it to act somewhere or you're never going to understand it.
557
00:53:23,350 --> 00:53:26,800
So we need mechanisms. If you give me some algebraic description of a group,
558
00:53:26,980 --> 00:53:33,070
I need to have some mechanism to conjure up some action on something so I can use geometry to try and understand it.
559
00:53:33,250 --> 00:53:40,000
And hopefully in doing this, I'm going to discover some nice new geometries. So let me skip ahead to this construction.
560
00:53:40,390 --> 00:53:45,820
So here's a group we do understand. We keep seeing this got two basic operations and they commute.
561
00:53:45,820 --> 00:53:47,170
Doesn't matter what order you do them in.
562
00:53:48,430 --> 00:53:57,130
Here is a purely geometric description without understanding that group where I can conjure up the sort of things that acts are so in
563
00:53:57,960 --> 00:54:05,260
I so often in mathematics you've got some data and it's come from some situation and now you want to extract more power out of that.
564
00:54:05,590 --> 00:54:10,290
So you sit back. I seem to be saying there's a lot of sitting back in mathematics, but there is.
565
00:54:10,870 --> 00:54:13,870
And you think, well, how else could I interpret that data?
566
00:54:14,020 --> 00:54:17,650
What else is naturally associated? That data that would change context.
567
00:54:17,860 --> 00:54:21,520
And maybe from that change perspective, I could actually get more information.
568
00:54:22,120 --> 00:54:28,840
So you look at this, at these symbols. Look at your notebook with the eyes of a topologies to a geometry instead of an algebraic.
569
00:54:29,740 --> 00:54:34,330
Okay. And I'll treat that as some meccano assembly instructions.
570
00:54:34,570 --> 00:54:42,130
And the meccano instructions are take one point a attach one wire for each of my generating objects,
571
00:54:42,910 --> 00:54:52,510
take a little rubber sheet and attach glue it and attach it to the wires according to spell out the word a B than a inverse and b inverse.
572
00:54:53,510 --> 00:54:56,660
If you do that, you'll get this bagel like object over here.
573
00:54:56,670 --> 00:55:01,490
It's called the Taurus. Okay. And then the magic of algebraic topology, which I'm not going to explain,
574
00:55:01,820 --> 00:55:06,530
is that if you're wandering around on here, imagine yourself a little ant wandering around on here.
575
00:55:07,430 --> 00:55:12,680
And imagine you're a member of the Flat Earth Society. You haven't yet learned to think globally.
576
00:55:12,830 --> 00:55:19,010
You're just really a little, aren't you wandering around on here? You don't know whether you're on here or you're on here.
577
00:55:20,230 --> 00:55:24,960
Right, because all you see is this little dot around you. If you're there, you've got two arrows.
578
00:55:25,230 --> 00:55:31,790
You've got to be arrows and all looks the same. So with a limited horizon, you don't know whether you live here or here.
579
00:55:32,750 --> 00:55:37,340
And so this unwrapping procedure gives me something where this group acts as symmetries.
580
00:55:37,700 --> 00:55:45,050
Now, I rather like that, right? Because I just took some systematic way of taking the symbols of the system of symmetry and conjuring up an action.
581
00:55:45,380 --> 00:55:51,890
I've got this group now acting on that thing up there. And so I'll take some less familiar a skip over this.
582
00:55:52,520 --> 00:55:57,620
I take some less familiar. So, hey, here's another. I do that with this group we had earlier.
583
00:55:58,130 --> 00:56:04,340
I do that little construction downstairs. Now I unwrap it, and upstairs in that picture, I'll find this nice tessellation.
584
00:56:05,600 --> 00:56:09,710
Now I'm about to get new geometry. Ready for new geometry. Observe.
585
00:56:09,750 --> 00:56:14,510
Here's something. Whose symmetries? Look my notes at B and C, you do them twice, you get nothing.
586
00:56:14,750 --> 00:56:19,490
And these numbers here, tell me what happens when you add a couple times two, three and six.
587
00:56:19,700 --> 00:56:23,880
See those numbers? Two, three and six. Let's make the least possible change.
588
00:56:23,900 --> 00:56:27,620
Let's turn two, three and six into two, four and six.
589
00:56:28,700 --> 00:56:32,870
And repeat that operation. Okay. Take that group presentation.
590
00:56:32,870 --> 00:56:37,219
With the minimal change that was previously given in a copy of the usual tiling at the plane,
591
00:56:37,220 --> 00:56:43,370
we're all used to go through that system of generating objects on which the group would be symmetries.
592
00:56:43,700 --> 00:56:48,200
And you got this wonderful object. Okay. Make another slight change.
593
00:56:48,200 --> 00:56:55,130
And you got this very similar looking object. And what's happened here is we've discovered something called hyperbolic geometry.
594
00:56:55,400 --> 00:56:58,400
Excuse me. So Euclidean geometry is the one we.
595
00:56:58,490 --> 00:57:04,190
We sort of naively used to have squares going off forever. Hyperbolic geometry is what's pictured here.
596
00:57:05,120 --> 00:57:09,529
And if you think of this as an infinite space that you think of, this is a road system.
597
00:57:09,530 --> 00:57:11,870
You've got to pay a pound to go along each edge.
598
00:57:12,700 --> 00:57:16,940
And so the distance between two points, how much you've got to pay to get from one point to the other.
599
00:57:17,780 --> 00:57:22,230
So even though these look like they're getting tiny to go from there to there, you've got to pay that.
600
00:57:22,250 --> 00:57:27,420
Not not not that. So is that the infinite? Because you got across infinitely many edges to get to the end of the disk.
601
00:57:28,340 --> 00:57:33,860
There's an infinite geometry, different to Euclidean geometry that we've discovered by first,
602
00:57:33,860 --> 00:57:37,490
starting with the system of symmetry and then asking, what is it? Symmetries of.
603
00:57:38,660 --> 00:57:44,210
Now this is a rather beautiful thing that has all sorts of properties that are different to Euclidean space.
604
00:57:44,280 --> 00:57:47,570
It's very well-studied and particularly has thin triangles.
605
00:57:47,610 --> 00:57:53,900
And if you're wandering around in this space, you wouldn't think you on a flat plane, you'd think you at the point of a saddle all the time.
606
00:57:54,200 --> 00:57:58,160
So locally, if you're anywhere in the space and you're trying to, you think you're in a saddle.
607
00:57:59,550 --> 00:58:07,610
But this really happens in nature, right? So what happens in hyperbolic geometry is things try to grow exponentially quickly.
608
00:58:07,740 --> 00:58:14,549
If you count the number of triangles within that, then PN steps of that point, you'll find it's growing exponentially.
609
00:58:14,550 --> 00:58:20,400
It's multiplying roughly by two at each step. Now for a while you can do that in Euclidean space.
610
00:58:20,700 --> 00:58:24,210
Vegetables try to do it and they end up looking curly like that.
611
00:58:24,820 --> 00:58:29,250
A coral tries to do it, ends up being negatively curved, hyperbolic like that.
612
00:58:29,490 --> 00:58:33,000
Even bizarrely, some crochet enthusiasts do it.
613
00:58:34,680 --> 00:58:38,970
Well, that's a lot to think about. And I just want to finish the couple of remarks.
614
00:58:39,180 --> 00:58:43,500
Oh, I should have stopped. So this hyperbolic geometry happens in every dimension.
615
00:58:43,950 --> 00:58:48,930
Okay. More than that, there's all sorts of other geometries I haven't told you about that come out of fact,
616
00:58:49,410 --> 00:58:51,480
that come out of thinking about arbitrary groups.
617
00:58:52,200 --> 00:59:00,120
So you go to think, you go away and you think about this and you realise that as I was standing there in the Isle of Man, I should have zoomed out.
618
00:59:01,020 --> 00:59:05,270
And when I assumed out. At the next step.
619
00:59:05,270 --> 00:59:08,330
I really can't be confident what I get. Right.
620
00:59:08,570 --> 00:59:16,130
There's lots of geometry, so maybe I have flat geometry or total geometry or spherical geometry or some hyperbolic geometry.
621
00:59:17,600 --> 00:59:19,639
Okay. So think about this.
622
00:59:19,640 --> 00:59:26,750
All the young people in the audience think about can you decide what sort of surface you live on by just making experiments?
623
00:59:26,780 --> 00:59:33,499
Can you do some ordnance survey just on the earth and figure out whether you live on a sphere or doughnut or flat
624
00:59:33,500 --> 00:59:39,080
plane or high agena surfaces that call something like that or something with three holes or four holes or more holes.
625
00:59:39,860 --> 00:59:44,809
Think of experiments. Can you do that? And it might not always be easy to recognise things.
626
00:59:44,810 --> 00:59:48,080
So that surface there is the same as that surface there.
627
00:59:49,430 --> 00:59:56,000
By which I mean if you cut all the tiles up there and reassemble them so they have the same edges in common, you'll get that previous surface.
628
00:59:57,080 --> 01:00:05,720
So again, it's hard to recognise which surface the same, but it's possible when you go into three dimensional space it's even harder.
629
01:00:05,780 --> 01:00:14,350
It takes most of 20th century mathematics. But you can make a complete list of all possible models of reasonable three dimensional spaces.
630
01:00:15,020 --> 01:00:19,909
The key people have PoincarĂ© who really developed that picture of hyperbolic geometry.
631
01:00:19,910 --> 01:00:25,490
I've explained to Bill first and who first thought it might be possible to classify space in three dimensions.
632
01:00:25,790 --> 01:00:28,790
And Gregory Fischer Perelman was the one who actually proved it.
633
01:00:29,330 --> 01:00:35,270
Okay, so having gone from 2 to 3 to 4 to 3 dimensions, should we go to four dimensions?
634
01:00:36,870 --> 01:00:43,050
No. Because just as it's impossible to tell what a system of symmetry is.
635
01:00:43,400 --> 01:00:47,240
Well, it's impossible to tell if an arbitrary diophantine equation has a solution or not.
636
01:00:47,720 --> 01:00:57,920
When you get to dimension four explicitly because all groups come into play in Dimension three, there's some restrictions I mentioned for all groups,
637
01:00:57,920 --> 01:01:05,300
all systems of symmetry you can never describe arise as symmetries of models of space time, so-called four dimensional manifold.
638
01:01:05,690 --> 01:01:11,030
And because you can't recognise what group you've got, you can't recognise which space time you've got.
639
01:01:11,720 --> 01:01:16,640
So once you can classify all models of two dimensional space and all modes of three dimensional space,
640
01:01:16,850 --> 01:01:24,980
it's proved impossible to make a proper, exhaustive, ever done list of all models of full space explicitly because of group theory.
641
01:01:25,790 --> 01:01:27,860
Okay, that's a good place to stop. Thank you.