Ruan's English + Hindi Log, written in English w/ Portuguese translation

Continue or start your personal language log here, including logs for challenge participants
Ruan
Yellow Belt
Posts: 99
Joined: Thu Aug 27, 2015 1:05 pm
Languages: Brazilian Portuguese (N), English
Language Log: viewtopic.php?f=15&t=1370
x 154

Ruan's English + Hindi Log, written in English w/ Portuguese translation

Postby Ruan » Fri Sep 25, 2015 9:31 am

September 25th

Hindi
[*] I'm trying to get past the Devaganari script brickwall. Right now that sums up to drilling letters by rewriting them over and over, five at a time. Cover five Devanagari letters, write five Latin letters. Write five Latin letters, cover five Devanagari letters. Wash, rinse, repeat until you're done.
[*] Estou tentando atravessar a barreira da escrita Devanagari. Agora isso se resume a decorar letras através da reescrita, várias vezes, cinco letras de cada vez. Cobrir cinco letras em Devanagari, escrever cinco letras romanas. Cobrir cinco letras romanas, escrever cinco letras em Devanagari. E por aí vai até terminar de decorar.

English
Image

[*] I've been memorizing vocabulary through this method. This is done on a little notebook. Left page, I write the words I come across I don't know yet, as columns. Right page, I write the columns as lines, four to seven words per line, and the translations below each word. Then I cover the first line with a paper sheet, and I rewrite the first from reading the second. Then I write the fourth line, which is the second from reading the third. This drilling goes on until I'm confident enough for step two.
[*] Estou memorizando vocabulário por esse método. Uso um caderninho. Página esquerda, escrevo as palavras que ainda não sei, como colunas. Página direita, escrevo as colunas como linhas, quatro a sete por linha, e as traduções debaixo de cada palavra. Então eu cubro a primeira linha com um pedaço de papel, e reescrevo a primeira a partir da segunda. Então eu escrevo a quarta linha, que é a segunda a partir da terceira. E por aí vai até eu ter confiança para partir para a segunda etapa.

Image

[*]This is step two: I flip the page and rewrite the left page columns as lines. Then, the next day I fill in the gaps by writing their translations from memory. If there's anything I can't remember, I drill it again.

You can notice that left pages have dates. That's because I'll review these pages every two months for a year. Again, if there's anything unknown, I'll go through the same process again.

I'm skeptical of spaced repetition being so much more effective than equal spaced repetitions over large periods of time, and I've read some research that backs up my intitution. So for now I'm sticking with this notebook method ; pen and paper feels more comfortable for me than SRS software, like Anki.

[*]Essa é a segunda etapa: viro a página e reescrevo as colunas da página esquerda como linhas. Então, no outro dia eu cubro as lacunas escrevendo as traduções de memória. Se tem algo de que não me lembro, eu decoro de novo.

Dá pra ver que as páginas esquerdas tem datas. Isso é porque eu vou revisar estas páginas a cada dois meses durante um ano. Assim, se tiver algo que não sei mais, eu repito o processo.

Eu não coloco muita fé em dizer que repetição espaçada é muito mais eficiente que repetições iguais durante largos períodos de tempo, e eu li algumas pesquisas que sustentam essa minha intuição. Por isso, por enquanto eu estou me aferrando a esse método do caderninho ; lápis e papel é bem mais confortável pra mim que programas de SRS, como o Anki.

[*] My main source of vocabulary is to google for blog posts from the last hour. I do this by typing "<random word> site:blogspot.com" and restricting the results to the last hour. This way I get input from a lot of different sources and I don't get stuck inside some writer's semantic bubble for long timespans.

[*] Minha principal fonte de vocabulário é procurar por posts em blogs da última hora no Google. Eu faço isso digitando "<palavra aleatória> site:blogspot.com" e restringindo os resultados para a última hora. Isso me dá muito material de várias fontes diferentes, e eu não fico preso dentro da bolha semântica de algum escritor por um lapso temporal muito largo.
Last edited by Ruan on Sun Sep 27, 2015 6:08 pm, edited 1 time in total.
1 x

Ruan
Yellow Belt
Posts: 99
Joined: Thu Aug 27, 2015 1:05 pm
Languages: Brazilian Portuguese (N), English
Language Log: viewtopic.php?f=15&t=1370
x 154

Re: Ruan's English + Hindi Log, written in English w/ Portuguese translation

Postby Ruan » Sat Sep 26, 2015 9:06 am

September 26th

Yesterday I've focused on developing a technique to improve my pronunciation. I've never consciously worked on my English pronunciation and people already say my accent is better than most of the Brazilians they've ever met. As all I've done was massive amounts of listening and little to no speaking pratice, I guess that's okay to postpone work on your accent until lhe later learning stages. Toddlers already do that; they won't talk until they've got one year or two of massive listening.

Ontem eu me enfoquei em desenvolver uma técnica para melhorar minha pronunciação. Eu nunca trabalhei de modo consciente na minha pronúncia do inglês, e as pessoas já dizem que meu sotaque é melhor do que a maioria dos brasileiros que eles já conheceram. Já que tudo que eu fiz foi escutar grandes quantidades de áudios e quase não treinar a fala em si, eu acho que pode ser bom adiar o trabalho sobre o sotaque até os estágios mais avançados. Os bebês já fazem isso; eles não falam até terem escutado muita linguagem por um ano ou dois.

I'm spending a lot of time listening to Jain bhajans and stavans for my Hindi; it's enjoyable and it trains my ear to the language at the same time. As for my English, I'll download some vlogs and podcasts from LA. Right now I'll focus on tomexploresla on YouTube. I'll do listen-pause-repeat until I memorize the vlog; after I'm done memorizing, I'll repeat at the same time as the vlogger ( so-called shadowing ). I'll repeat just around twenty minutes of recordings for several times, maybe even during a whole month or more. Then I'll post 10-second recordings of my attempt on lang-8 and ask for advice. I'm doing this with vlogs because I'm at a stage where I can understand pretty much anything natives say; that's why I won't be shadowing Hindi vlogs any time soon. For Hindi, I guess I'll rather start by memorizing songs and poetry when I feel I'm ready.

Eu estou gastando muito tempo escutando Jain bhajans e stavans pro meu Hindi; é agradável e treina meu ouvido pra língua ao mesmo tempo. Pro meu inglês, eu vou baixar alguns vlogs e podcasts de Los Angeles. Agora vou me focar em tomexploresla no YouTube. Eu escuto-pauso-repito até memorizar o vlog; depois de ter memorizado, eu repito ao mesmo tempo que o vlogger ( o famoso shadowing ). Eu repito por volta de vinte minutos de gravações por várias vezes, talvez durante um mês inteiro ou mais. Então eu posto gravações de 10 segundos da minha tentativa no lang-8 e peço por conselhos. Estou fazendo isso com vlogs porque no meu estágio atual eu consigo entender praticamente qualquer coisa que os nativos falam; e é por isso que eu não vou fazer shadowing em cima de vloggers de Hindi tão cedo. Pro Hindi, ao invés disso, vou começar memorizando poesias e músicas quando eu me sentir pronto para isso.
0 x

Ruan
Yellow Belt
Posts: 99
Joined: Thu Aug 27, 2015 1:05 pm
Languages: Brazilian Portuguese (N), English
Language Log: viewtopic.php?f=15&t=1370
x 154

Re: Ruan's English + Hindi Log, written in English w/ Portuguese translation

Postby Ruan » Sun Sep 27, 2015 10:39 am

Today I've set up Firefoxto open a random English word from the Wiktionary every time I open a new tab by pressing ctrl + t. Then I find blog articles to read by searching for "word site:blogspot.com".

I really want to take a break from the screen, so I guess I'll print a whole bunch of blogspot articles in a row for later reading. So far, here's my routine for English:

Pronunciation practice: Listen-and-repeat, and then shadowing, on 20 minutes of audio downloaded from tomexploresla.

Reading pratice: Reading the aforementioned blogs.

Listening practice: I'll listen to the entire Santa Barbara corpus. I've downloaded everything with wget and it's 23 hours long. I've also tried to make the .trn files more printer-friendly, see here.

Hoje eu ajustei o Firefox para abrir uma página aleatória do Wiktionary toda vez que eu abro uma nova aba ao apertar ctrl +t. Aí eu acho artigos de vlogs para ler buscando por "palavra site:blogspot.com".

Eu quero mesmo é dar um tempo das telas, então acho que vou imprimir um bom tanto de artigos do blogspot de uma vez para ler depois. Até agora, essa é a minha rotina pro inglês:

Prática de pronúncia: Escutar e repetir, e então shadowing, em 20 minutos de áudio baixado do tomexploresla.

Prática de leitura: Ler os blogs supracitados.

Prática de escuta: Vou escutar todo o Santa Barbara corpus. Eu baixei tudo com wget e são 23 horas de duração. Tentei fazer os arquivos .trn ficarem mais imprimíveis, veja aqui.
Last edited by Ruan on Mon Sep 28, 2015 1:15 pm, edited 2 times in total.
0 x

Ruan
Yellow Belt
Posts: 99
Joined: Thu Aug 27, 2015 1:05 pm
Languages: Brazilian Portuguese (N), English
Language Log: viewtopic.php?f=15&t=1370
x 154

Re: Ruan's English + Hindi Log, written in English w/ Portuguese translation

Postby Ruan » Sun Sep 27, 2015 4:41 pm

This script downloads all blogspot posts from a link list in a text file and puts them in a printer-friendly format.
Esse script baixa todas postagens no blogspot de uma lista de links em um arquivo de texto e as coloca em um formato mais imprimível.

Code: Select all

#! /bin/bash
COUNTER=0
while read curline; do
   COUNTER=$((COUNTER+1))   
   lynx -dump "$curline" > "$COUNTER.lynx.txt"
done < list.txt

for file in *.lynx.txt
do
   sed -i '/ Posted /I,$ d' $file
   sed -i '/ Comment /I,$ d' $file
   sed -i '/ Archive /I,$ d' $file
   sed -i '/Visible links/I,$ d' $file
   sed -i '/\* \[/I,$ d' $file
   fmt -w 170 $file > $file.keep
   rm $file
   mv $file.keep $file
done
cat *.lynx.txt > result.txt
rm *.lynx.txt



Output example.
Last edited by Ruan on Sun Sep 27, 2015 8:39 pm, edited 1 time in total.
0 x

Ruan
Yellow Belt
Posts: 99
Joined: Thu Aug 27, 2015 1:05 pm
Languages: Brazilian Portuguese (N), English
Language Log: viewtopic.php?f=15&t=1370
x 154

Re: Ruan's English + Hindi Log, written in English w/ Portuguese translation

Postby Ruan » Sun Sep 27, 2015 5:52 pm

Repost from another thread.

As I've already described in my log, here's my present routine for English:

[*]Listening to the Santa Barbara Corpus of Spoken American English - I'll listen to the entire corpus. It's 23 hours long, and I'll listen each file more than once, so I guess it'll keep me busy from fifteen days up to a month. This thing is a goldmine, it's like eavesdropping natives over the internet. I wish people were doing the same for the other languages as well.

[*]Shadowing videos from a YouTube channel called tomexploresla. It's a vlog where pratically everybody that appears on screen was born and raised in LA, so I'm using this to narrow my accent down to a specific dialect. I'll listen-pause-repeat ( Pimsleur-like ) to a very small set of recordings taking from this channel ( twenty minutes long or so ). When I feel I pretty much know them by heart, then and only then I'll start shadowing. The entire process will surely take more than a month.

[*]Reading randomly chosen blog articles from blogspot.com. I won't post many details here ( you can read them on my log ) but I'm reading blogs that come up when I google for "word search:blogspot.com", in which "word" is taken from Wiktionary:Random_page or random words from the Online Slang Dictionary.

[*]Things that are not "routine" but that I do anyway are talking to random people on IRC and asking for my lang-8 fellows to proofread my writing or to judge my accent.

My Hindi routine, on the other hand, is much more simpler, as I'm still taking baby steps. For now it boils down to memorizing Devanagari and basic words, and listening to Jain music.

At least for me, a key thing to get myself concentrated is to turn off the screens and work with actual printed paper. That's why I even wrote a script to batch print blog posts from a link list.
0 x

Ruan
Yellow Belt
Posts: 99
Joined: Thu Aug 27, 2015 1:05 pm
Languages: Brazilian Portuguese (N), English
Language Log: viewtopic.php?f=15&t=1370
x 154

Re: Ruan's English + Hindi Log, written in English w/ Portuguese translation

Postby Ruan » Tue Sep 29, 2015 12:26 am

As I'm striving for real trilingualism, I'll often expose myself to randomly selected materials, spread throughout distinct linguistic environments, instead of taking the typical polyglot approach of reading about whatever narrow range of subjects that one finds most appealing.

COCA, the Corpus of Contemporary American English, uses five resources to build up a balanced vocabulary database: transcripts of unscripted conversation from television and radio, novels and plays, magazines, newspapers and academic journals.

All but the first are easy to get. Spoken language recordings and their transcriptions are ideally available through corpora; if no corpus is available for the language, then one will have to resort to blog comments, reddit posts, songs and their lyrics, tweets, IRC logs, and any other sources that come close to spoken language. Blog posts and comments are easier to fetch and format with standard unix tools, and there's an option to search for blogspot blogs by location.

It's possible to mine unscripted spoken language from video sharing websites, television and radio; however, it's almost always fated to be too time consuming, as our goal is to spend as little time as possible in gathering phase. Most videos posted on video sharing websites sound scripted, as their authors intended to perform rather than to talk freely.
0 x


Return to “Language logs”

Who is online

Users browsing this forum: Cavesa, guyome, Iversen and 2 guests