DUBLIN- text in public space
TEXT IN PUBLIC SPACE - a message decrypted
An attempt to use a scientific data for art creation. Project for future painting.
In this project, I have combined translation studies with artistic investigations and measured the energy of places by corpus based analysis of semantic prosody in texts embedded in the public space.
People are surrounded by a variety of texts. Their sense is clear, but sometimes they convey additional implications regarding the reception of a message. Here, I would like to pertain to semantic prosody. The term describes a phenomenon where certain seemingly neutral words can be perceived with positive or negative associations through frequent occurrences with particular collocations. In other words, some words tend to go together and inherit from each other negative or positive connotations. Using computational linguistics tools, I measured the energy of places. Currently, I am concentrating on big cities with a large amount of textual data.
I took photos of texts embedded in public space, then established the frequency of words with a computer program, and, finally, investigated semantic prosody of the most frequent nouns and verbs. These hidden energies are represented by the colours of my choice, and their consistent use would create a map of a city.
DUBLIN
Wordlist: the top 20 most frequent words
1 # 205 10.09354973
2 THE 64 3.151157141
3 TO 31 1.526341677
4 APRIL 24 1.181683898
5 MON 20 0.984736562
6 THEATRE 20 0.984736562
7 SAT 19 0.935499728
8 DUBLIN 18 0.886262953
9 S 18 0.886262953
10 AND 17 0.837026119
11 MAY 16 0.787789285
12 OLYMPIA 14 0.689315617
13 FOR 13 0.640078783
14 OF 13 0.640078783
15 A 12 0.590841949
16 IE 12 0.590841949
17 LET 12 0.590841949
18 SQ 12 0.590841949
19 ON 9 0.443131477
20 YOU 9 0.443131477
Disregarding function words and proper nouns/names, the following words have been distinguished:
Let, sq, you.
Subsequently, candidate collocations were computed in Sketch Engine using BNC (British National Corpora)
1.1. 'You' collocation candidates.
1.2. 'Let' collocation candidates.
1.3. 'Sq' collocation candidates.
Although all measurements were taken in accordance with the rules established in the scientific research, the size of sample unables to treat it as a credible evidence on which can be made generalisation of findings.
The choice of visual representation and colours was based on personal preferences. This will not be further explained. The most difficult was to find colours for representation of negative prosody. None of the colours is negative, however, some of them can be perceived as dirty, not stick to each other, and so on. The conclusion on data evaluation is left to the observers/recipients of the visuals. As for me, the connotations of 'you' are manipulative (you do not, you do). 'Let' has a negative prosody (refused, refusing). Even though the results may pertain to a different context, the fragmentary perception of the street advertisements makes an issue of context unimportant. 'SQ' goes the most frequently with words such as population, km, DATA, BASIC. Thus, the omnipresence of the word SQ keeps us aware that the city along with its life and people is to be meticulously measured and evaluated.
Now I know why I come home from Dublin with such a bad headache!
An attempt to use a scientific data for art creation. Project for future painting.
In this project, I have combined translation studies with artistic investigations and measured the energy of places by corpus based analysis of semantic prosody in texts embedded in the public space.
People are surrounded by a variety of texts. Their sense is clear, but sometimes they convey additional implications regarding the reception of a message. Here, I would like to pertain to semantic prosody. The term describes a phenomenon where certain seemingly neutral words can be perceived with positive or negative associations through frequent occurrences with particular collocations. In other words, some words tend to go together and inherit from each other negative or positive connotations. Using computational linguistics tools, I measured the energy of places. Currently, I am concentrating on big cities with a large amount of textual data.
I took photos of texts embedded in public space, then established the frequency of words with a computer program, and, finally, investigated semantic prosody of the most frequent nouns and verbs. These hidden energies are represented by the colours of my choice, and their consistent use would create a map of a city.
DUBLIN
Wordlist: the top 20 most frequent words
1 # 205 10.09354973
2 THE 64 3.151157141
3 TO 31 1.526341677
4 APRIL 24 1.181683898
5 MON 20 0.984736562
6 THEATRE 20 0.984736562
7 SAT 19 0.935499728
8 DUBLIN 18 0.886262953
9 S 18 0.886262953
10 AND 17 0.837026119
11 MAY 16 0.787789285
12 OLYMPIA 14 0.689315617
13 FOR 13 0.640078783
14 OF 13 0.640078783
15 A 12 0.590841949
16 IE 12 0.590841949
17 LET 12 0.590841949
18 SQ 12 0.590841949
19 ON 9 0.443131477
20 YOU 9 0.443131477
Disregarding function words and proper nouns/names, the following words have been distinguished:
Let, sq, you.
Subsequently, candidate collocations were computed in Sketch Engine using BNC (British National Corpora)
1.1. 'You' collocation candidates.
1.2. 'Let' collocation candidates.
1.3. 'Sq' collocation candidates.
Although all measurements were taken in accordance with the rules established in the scientific research, the size of sample unables to treat it as a credible evidence on which can be made generalisation of findings.
The choice of visual representation and colours was based on personal preferences. This will not be further explained. The most difficult was to find colours for representation of negative prosody. None of the colours is negative, however, some of them can be perceived as dirty, not stick to each other, and so on. The conclusion on data evaluation is left to the observers/recipients of the visuals. As for me, the connotations of 'you' are manipulative (you do not, you do). 'Let' has a negative prosody (refused, refusing). Even though the results may pertain to a different context, the fragmentary perception of the street advertisements makes an issue of context unimportant. 'SQ' goes the most frequently with words such as population, km, DATA, BASIC. Thus, the omnipresence of the word SQ keeps us aware that the city along with its life and people is to be meticulously measured and evaluated.
Now I know why I come home from Dublin with such a bad headache!
Comments 0
Say something