WEBVTT
Kind: captions
Language: en-GB

1
00:00:00.089 --> 00:00:04.320
The smallest unit of meaning in a language
is called a “morpheme”. Let’s take a

2
00:00:04.320 --> 00:00:09.130
lovely English word, “inconceivable”.
There are three morphemes there:

3
00:00:09.130 --> 00:00:13.510
conceive, which in this context means to form
something in your head. Now you could break

4
00:00:13.510 --> 00:00:18.360
that down further if we were in old Latin,
but in English that is a morpheme: con and

5
00:00:18.360 --> 00:00:20.910
ceive don’t mean anything on their own here.

6
00:00:20.910 --> 00:00:26.240
Then we’ve got -able, which means the possibility
of something. “Conceivable”. We can think

7
00:00:26.240 --> 00:00:27.699
about this.

8
00:00:27.699 --> 00:00:33.989
And then we've got in-, which negates it.
“Inconceivable”. We cannot possibly think

9
00:00:35.660 --> 00:00:39.730
Different languages have very different approaches
on how to deal with morphemes, and those approaches

10
00:00:39.730 --> 00:00:44.160
are the reason why some languages have many
short words, and others have long structures

11
00:00:44.160 --> 00:00:47.800
that are frequently difficult for adults learning
a new language to deal with.

12
00:00:47.800 --> 00:00:53.329
There’s a spectrum, and it stretches from
“analytic” to “synthetic”.

13
00:00:53.329 --> 00:00:57.789
Over on the analytical side, there are the
isolating languages, like Chinese and Vietnamese.

14
00:00:57.789 --> 00:01:03.149
Here, each morpheme is usually an entirely
separate word. Assembling a sentence in one

15
00:01:03.149 --> 00:01:07.590
of these languages means you’re mostly picking
and choosing words and putting them next to

16
00:01:07.590 --> 00:01:08.069
each other.

17
00:01:08.069 --> 00:01:13.270
Next along this spectrum, there are the agglutinative
languages, like Turkish and Inuktitut. “Agglutinative”

18
00:01:13.270 --> 00:01:18.459
- it's a lovely word - has the same Latin
roots as the word “glue”: you’re gluing

19
00:01:18.459 --> 00:01:23.209
words together from their component parts.
Rather than picking extra words to add to

20
00:01:23.209 --> 00:01:27.950
your sentence, you’re adding prefixes or
suffixes to words that are already there.

21
00:01:27.950 --> 00:01:32.929
Depending on the language, you might add affixes
for tense, person, number, belonging, possession,

22
00:01:32.929 --> 00:01:36.399
or even things like whether an action was
deliberate or not.

23
00:01:36.399 --> 00:01:40.659
Next up are the fusional languages. They work
in the same way -- assembling bits to make

24
00:01:40.659 --> 00:01:45.700
a full word with all the meaning you want
-- but now the bits you’re adding affect

25
00:01:45.700 --> 00:01:50.569
the parts you’ve got already, tweaking how
it looks or how it sounds. Not only that,

26
00:01:50.569 --> 00:01:54.669
but each of the bits you’re adding can have
multiple different meanings attached to it:

27
00:01:54.669 --> 00:01:58.749
tense, number, person, all sorts of things
can be coded with just one sound attached

28
00:01:58.749 --> 00:02:05.349
to a word. Take “hablo” in Spanish. That
-o morpheme? It means first person, singular,

29
00:02:05.349 --> 00:02:11.620
present, indicative. That’s a lot of meaning
in a very short sound.

30
00:02:11.620 --> 00:02:19.379
And then, all the way over here, are the polysynthetic
languages, basically the really, really extra-synthetic

31
00:02:19.379 --> 00:02:24.480
languages. You’ll find this in Algonquian
languages, up in the cold parts of North America.

32
00:02:24.480 --> 00:02:28.930
This is where you can combine potentially
a dozen or more morphemes, enough for a whole

33
00:02:28.930 --> 00:02:35.379
sentence, a whole coherent thought, into one
long word. And it really is a word: the individual

34
00:02:35.379 --> 00:02:39.310
parts can’t be split apart and survive on
their own. The morphemes you’re using might

35
00:02:39.310 --> 00:02:43.799
be agglutinative, in separate blocks, or they
might be fusional, affecting everything around

36
00:02:43.799 --> 00:02:51.909
them, but at any rate you have an entire sentence
in the form of one beautiful, long, interconnected

37
00:02:51.909 --> 00:02:52.430
word.

38
00:02:52.430 --> 00:02:57.549
Now here’s the thing: most English speakers
consider polysynthetic languages to be the

39
00:02:57.549 --> 00:03:02.689
crazy side of language. It seems incredibly
complicated. And for adults trying to learn

40
00:03:02.689 --> 00:03:07.749
a polysynthetic language, it is incredibly
confusing. But children may actually find

41
00:03:07.749 --> 00:03:12.640
these languages easier to learn: because the
words are long and intertwined, and each bit

42
00:03:12.640 --> 00:03:17.340
affects another, there’s more redundancy
there. If you mishear or misunderstand one

43
00:03:17.340 --> 00:03:21.659
part, you’re much more likely to be able
to work out what is meant from how the words

44
00:03:21.659 --> 00:03:22.930
around it have changed.

45
00:03:22.930 --> 00:03:27.129
It’s important to remember that like many
things in life, this is a spectrum. There

46
00:03:27.129 --> 00:03:31.519
aren’t many languages that fit neatly into
any one of these categories. English is actually

47
00:03:31.519 --> 00:03:35.639
a fairly isolating language, down at this
end of the scale -- but it’s still generally

48
00:03:35.639 --> 00:03:39.799
classified as synthetic, because we’ve got
many, many words like “inconceivable”.

49
00:03:39.799 --> 00:03:44.700
And the family that English belongs to, Indo-European,
is pretty synthetic, too.

50
00:03:44.700 --> 00:03:48.909
And while this is the conventional approach
to classifying languages, it does have a few

51
00:03:48.909 --> 00:03:53.680
problems. It distinguishes loads of different
categories on this side of the spectrum, but

52
00:03:53.680 --> 00:03:57.810
it lumps all the isolating languages together.
That’s probably because most of the linguists

53
00:03:57.810 --> 00:04:03.150
who set up the classification -- particularly
the early ones -- were European, so they concentrated

54
00:04:03.150 --> 00:04:06.299
on the differences that made sense to them.

55
00:04:06.299 --> 00:04:11.819
Languages are messy things that shift and
change over time, between places and people.

56
00:04:11.819 --> 00:04:15.709
Categorising them can be useful for research,
and it tells you a few things about the possibilities

57
00:04:15.709 --> 00:04:21.090
of what the human brain can do, but once you
start to file them away you start to realise

58
00:04:21.090 --> 00:04:26.110
that, like most everything about human communication,
they really don’t fit into neat little boxes.

59
00:04:27.960 --> 00:04:29.960
[Translating these subtitles? Add your name here!]