Panini Senior Linguistics Olympiad 2014
Round 1, 5th January 2014
Contest Problems
Maximum Marks: 100 Total Time: 4 hours
Instructions:
1. There are 8 pages in this question booklet, including this page. Please check and report to
the invigilator immediately if your booklet has less than 8 pages.
2. There are 6 problems. The total marks for each problem is indicated in the parenthesis next
to the problem title. A problem can have several sub-parts.
3. Additional information and interesting facts related to the problem are presented in the
gray boxes. While the problems can be solved without the help of this additional
information, the gray boxes might provide useful hints.
4. Write down your solution to each problem on a separate sheet or sheets. On each sheet
indicate the number of the problem, your seat number and your name. Otherwise your
work may be mislaid or misattributed.
5. Do not copy the statements of the problems.
6. All answers must be well-argumented. Even a perfectly correct answer will be given a low
score unless accompanied by an explanation.
7. Use of calculators, mobile phones and any other electronic devices is strictly prohibited. No
books, notebooks and other printed materials can be consulted during the contest.
Good Luck!
Problem Design Board:
Kalika Bali, Monojit Choudhury, Asif Ekbal, Shireen Irani, Mitesh Khapra,
Shakuntala Mahanta, Prachatos Mitra, Nilai Sarda, Zarana Sarda, Yash Sinha
1
Problem #1: Deciphering the Brahmi Script (15 marks)
An archeologist found a rock inscription from the 3rd century BC created by the Great Mauryan
Emperor Ashoka. She examined the first five lines of this Ashokan edict, reproduced below:
She immediately figured out that it is in Prakrit – an ancient Indian language and is written using the
Brahmi script. By comparing with other Brahmi inscriptions from the same time period, the
archeologist could decipher the first line of this edict, which is:
devānaṁpiye piyadasi lājā hevaṁ āhā ye atikaṁtaṁ
The archeologist could not decipher the other lines completely, but she made a list of seven Prakrit
words which she believed were present in the edict:
etaṁ, pari, nocujane, ca, esame, ichisu, dhaṁmavaḍhiyā
However, later it was found out that one of these seven words is actually not present in these five
lines.
Assignment 1:
A line over a vowel indicates a long vowel. A
Find the word that is not present.
dot under ṭ or ḍ indicates a retroflex sound
(like in Hindi ṭamāṭar and ḍamru), and is Assignment 2:
different from t and d without a dot, which Match the remaining words from the list to
indicates a dental sound (like in Hindi tum and the words in the Brahmi text.
desh). The letter ṁ indicates nasal sound.
Assignment 3:
Brahmi was an ancient Indian script. It most
Write down the following Prakrit words in
famously features in edicts of the Mauryan
Emperor Ashoka. In modern times, it was English using the transcription scheme used
deciphered by James Prinsep, a British here (see the gray box for details of
archaeologist, in 1837. transcription).
– Yash Sinha
2
Problem #2: Telling Time in Swahili (15 marks)
Here are eight time expressions in English and their unordered Swahili translations.
Sunday, 1:00 am jumamosi, saa moja usiku
Sunday, 7:30 am jumapili, saa tatu na robu asubuhi
Sunday, 9:15 am jumamosi, saa saba usiku
Tuesday, 12:15 pm jumamosi, saa mbili na robu usiku
Tuesday, 11:30 pm jumanne, saa tano na nusu usiku
Saturday, 10:30 am jumanne, saa sita na robu asubuhi
Saturday, 7:00 pm jumamosi, saa nne na nusu asubuhi
jumapili, saa moja na nusu asubuhi
Saturday, 8:15 pm
Assignment 1: Assignment 3:
Match the translations. Justify your solution. Translate the following English times into
Swahili
Monday, 12:15 am;
Assignment 2: Monday, 10:00 pm
Translate the following Swahili times into The Swahili language, or Kiswahili, is a Bantu
language spoken by various communities
English. inhabiting Tanzania, Kenya, Uganda, Rwanda,
jumatano, saa moja na robu asubuhi; Burundi, Mozambique and the Democratic
Republic of Congo. Around five million people
jumapili, saa nne na nusu asubuhi
speak Swahili as their mother tongue.
– Nilai Sarda
Problem #3: Bow or Onion? (15 marks)
Words often have multiple meanings. For example, the English word crane can either mean ‘a bird’
or `a machine used for lifting heavy objects`. Similarly, the Russian word лук (pronounced as luk)
has two meanings: onion and bow (a weapon). In the first paragraph below the word luk means
onion whereas in the second paragraph it means bow.
shalot i luk (onion) ochen tesno svjazannye ovoschi. oba imejut lukovichnye
rastenija. v vkusa, shalot bolee ostryj i, sledovatelno, ne edjat v kolichestvakh.
bak gorazdo bolee prodvinutyj, oruzie, chem luk (bow). vmesto togo, chtoby streljali
iz lukov ona ispolzuetsja, chtoby streljat snarjadami. eti snarjady ispolzujutsja v
sovremennoj vojne, chtoby ubit vragov.
3
For each of the following paragraphs containing the word luk, identify whether it means onion or
bow? Explain how you arrived at your conclusions. For ease of reading, the Cyrillic letters used for
writing Russian have been mapped to similar sounding Latin letters.
1. pistolet javljaetsja gorazdo bolee sovremennyj oruzie, chem luk. vmesto togo, chtoby strely iz
lukov on ispolzuetsja dlja metanija puljami. on ispolzuetsja v sovremennoj okhoty ubit zivotnykh.
2. chesnok i luk tesno svjazannye ovoschi . oni oba imejut lukovichnye rastenie. v vkusa, chesnok
bolee ostryj i, sledovatelno, ne edjat v bolshikh kolichestvakh.
3. okhotnikov ispolzovat luk iz dereva dlja okhoty na razlichnykh zivotnykh. oni chasto imel navyki
v strelbe iz luka i praktika ego s samogo rannego vozrasta. oni ispolzujut luk, metanija strely i
ubit nichego ne podozrevajuschikh zertv.
4. luk luchshe kultiviruetsja v plodorodnykh pochvakh, kotorye khorosho drenirovannye. v otlichie
ot mnogikh drugikh ovoschej, luk trebujut vysokogo urovnja pitatelnykh veschestv v pochve.
rastenie trebuet 11-12 chasov solnechnogo sveta.
5. luk soderzit paru izognutykh uprugikh konechnostej iz dereva. obe konechnosti soedineny stroki.
okhotniki specialistam v strelbe iz luka tjanet stroku nazad, chtoby metanija.
6. luk ispolzuetsja kak ovosch i javljaetsja naibolee shiroko kultiviruemykh i edjat vidov. eto
naibolee chasto raz v dva goda ili mnogoletnee rastenie. ego listja golubovato - zelenogo cveta.
7. luk i chesnok obychno ispolzujutsja v kachestve prigotovlenija pischi ingredienta dlja
povyshenija vkus indijskikh karri. on ispolzuetsja v kachestve zagustitelja dlja karri. on takze
edjat kak salat vmeste s drugimi ovoschami.
8. luki i strely byli shiroko ispolzuetsja v primitivnoj vojny. v sovremennoj vojne, oni byli
zameneny orudij i raket. khotja strelba iz luka bolshe ne trebuetsja v sovremennoj vojne ne do
sikh por praktika kak vid sporta.
9. luk i strely predstavljaet soboj sistemu snarjad oruzie. strelba iz luka eto iskusstvo, praktika ili
iskusstvo ispolzovanii luka dlja metanija strel. eto ochen primitivnoj sistema ispolzuetsja
okhotnikov i peschernogo cheloveka.
10. luk marinovannyj v uksuse edjat v kachestve zakuski. ego listja inogda ispolzuet dlja podgotovki
indijskikh karri. luk takze mozno upotrebljat v pischu syrye otlichie chesnok, kotoryj takze
javljaetsja ovosch prinadlezaschikh k odnomu vidu.
Choosing the correct meaning of a word in a given context (sentence or paragraph) is an
important task which must be solved for developing language processing and understanding
software such as automatic translation systems. This task is known as Word sense
disambiguation. The cues that a computer uses to solve this task is probably similar to what you
will be using to solve this problem!
- Mitesh Khapra
4
Problem #4: Making Sense of Turkish Words (15 marks)
Given below are some words in Turkish and their English translations, in no particular order.
gözlemci, döndürmek, gündöndü, gözlükcü, şarkici,
çocukluk, gözlemek, pazar, pazartesi, cumartesi, güneşli
English translations:
Saturday, Sunday, Monday, observer, singer
to observe, to rotate, sunny, sunflower, optician, childhood
Assignment 1: Hint: Read all parts of the question carefully,
Match the Turkish words with their English before working out the solution.
translations.
Turkish belongs to the Turkic language family.
Assignment 2: It is spoken in Turkey and within smaller
Translate the following words into Turkish: communities in many other countries. It is the
observation, child, the state of mother tongue of more than 63 million
being a singer, spectacles, Friday people. g, d, p, t, c, s, z, m, k, l, ş, ç denote
consonants, and ö, e, i, ü, a are vowels.
– Monojit Choudhury
Problem #5: Japanese Compound Words (20 marks)
In Japanese, certain root words undergo a change at their left edge when they form compound
words under specific circumstances. This phenomenon is traditionally called Rendaku. Examine the
following examples of Japanese compound words:
Japanese words Compound word Gloss Meaning
hito + hito hitobito person + person people
ike + hana ikebana keep alive + flower flower arrangement
toki + toki tokidoki time + time sometimes
te + kami tegami hand + paper letter
hitori + tabi hitoritabi one person + travel travelling alone
tsuno + tokage tsunotokage horn + lizard horned lizard
yama + kawa yamakawa mountains + rivers mountains and rivers
aka + huda akahuda red + tag red tag
maki + susi makizusi rolling + sushi# rolled sushi
yama + kawa yamagawa mountain + river mountain’s river
hebi + tosi hebidosi snake + year snake year
eda + ha edaha branches + leaves branches and leaves
oo + kaze ookaze big + wind big wind
kokoro + tsukai kokorodzukai heart + using anxiety
denki + kurage denkikurage electric + jellyfish electric jellyfish
yuki + kaeri yukikaeri coming + going coming and going
hana + sono hanazono flower + garden flower garden
nise + karasu nisegarasu fake + crow fake crow
5
Assignment 1:
Fill in the following compound words.
Japanese words Compound word Gloss Meaning
wa + kasi ____________ Japanese + cake Japanese cake
yama + kazi ____________ mountain + fire forest fire
juzu + tsunagi ____________ rosary + sequence roping together
yama + tera ____________ mountain + temple mountain temple
ao + kawazu ____________ green + frog green frog
uwa + huta ____________ top + lid top lid
e + tako ____________ picture + kite picture kite
Assignment 2:
When compounds are made up of more than two words, things become truly interesting. Fill in the
blank entries (A to E) in the following Table:
Japanese words Compound Word Gloss Meaning
hosi-kaki-tsukuri hosi-gaki-dzukuri dried persimmon* making making dried
persimmon
nise-kami-tana nise-kami-dana fake divine altar a divine altar that is
fake
nise-kami-tana nise-gami-dana fake divine altar an altar for a fake
God
ori-kami-tana ori-gami-tana fold paper shelf origami shelf
nise-sakura-dayori A. ? fake cherry-blossom tidings tidings of fake
cherry-blossoms
nise-sakura-dayori nise-sakura-dayori fake cherry-blossom tidings fake tidings of
cherry-blossoms
nuri-hasi-hako B. ? lacquered chopstick box a box for lacquered
chopsticks
nuri-hasi-hako nuri-hasi-bako lacquered chopstick box a lacquered box for
chopsticks
nise-tanuki-siru C. ? pseudo tanuki** soup soup made from
pseudo tanuki
nise-tanuki-siru D. ? pseudo tanuki soup tanuki soup that is
not authentic
oya-ko-kenka E. ? parent child quarrel parent and child
quarrel
Japanese is spoken by about 125 million speakers primarily in Japan.
#
sushi is cooked vinegared rice combined usually with raw fish.
*persimmon is a fruit.
**tanuki is Japanese raccoon dog.
- Zarana Sarda
6
Problem #6: Encoding Trees (20 marks)
Linguistic structures, such as grammatical relations between words in a sentence, are often
represented as a Tree. For example, the following parse or phrase-structure trees represent two
different interpretations of the same sentence.
Tree-A Tree-B
The above trees are known as strictly binary trees because at every branching point of the tree,
there are exactly two branches – a left and a right one. Since drawing a tree takes more space and
time, trees are often encoded as strings. According to one such encoding scheme for binary trees,
the above trees are represented as:
Tree-A: I 3 shot 1 an 0 elephant 2 in 1 my 0 pajama
Tree-B: I 4 shot 3 an 0 elephant 2 in 1 my 0 pajama
For this encoding scheme, every distinct binary tree has a unique string representation and for every
distinct valid string representation (like the two examples above), one can also uniquely construct a
binary tree. Your task is to figure out this encoding scheme and prove certain facts about it.
Assignment 1:
Shown below are two phrase-structure trees, Tree-C and Tree-D for two different interpretations of
a sentence. Write down the string encodings for Tree-C and Tree-D. Explain your assumptions.
Tree-C Tree-D
Assignment 2:
Construct the trees corresponding to the following encoded representations (words have been
replaced by letters): a 0 b 3 c 0 d 1 e 2 f 0 g a0b1c2d3e4f
Assignment 3:
The following encodings do not represent binary trees. Convert them into valid tree encodings by
incrementing or decrementing only one of the numbers by exactly 1.
a0b2c0d2e1f0g a1b3c2d1e0f
7
There are several constraints that a string must satisfy to be a valid encoding for a binary tree. What
constraints do these strings violate, which makes them invalid?
Assignment 4:
Let a1 x1 a2 x2 a3 … an-1 xn-1 an be a valid representation of a binary tree for a n word sentence a1 a2
a3 … an-1 an where x1, x2, … xn-1 are the numbers separating the words. Prove that
𝑛−1
𝑛
2 − 1 ≥ ∑ 2𝑥𝑖 ≥ ⌊log 2 𝑛⌋ ⌊ ⌋
𝑛
2
𝑖=1
- Monojit Choudhury
-------------------------------------------- END OF THE BOOKLET ---------------------------------------
