譯文大賞

When Did Charts Become Popular?

Modern newspapers are overflowing with charts. The New York Times, Washington Post and Wall Street Journal dedicate immense resources to making graphics for both their print and online editions.

現代的報紙都必備各式各樣的圖表，像是《紐約時報》、《華盛頓郵報》和《華爾街日報》，無論是在實體報紙和網路上，都不約而同利用圖表傳遞它們想要傳達的信息。

But it was not always so.

但這樣的現象並不是自古以來就是如此。

The New York Times was first published in 1851. From nearly the beginning of its existence, the Times brimmed with data like stock prices, weather statistics, and sports scores. But for the first century of publication, the use of charts to display that data was rare.

紐約時報是在1851年創立的一間報社公司。自創立初期，紐約時報開始就刊登了許多資料，像是股價、天氣相關數據和體育賽事的比數。但創立後的近一百年，其實報社幾乎沒有使用圖表來呈現這樣的資訊。

In order to understand when charts started appearing in papers, we collected data on the number of charts in editions of the New York Times over its history. We checked a weekday, September edition of the paper every five years from 1855 to 2015. We chose the Times because of its rich archives and because we found it to be representative of trends in the newspaper industry.

那麼圖表何時開始出現在報紙上呢？為了找到答案，我們蒐集了從創辦以來所有的紐約時報，為了節省過多的找尋時間，我們選取其中每五年的九月號。那麼相信您會想問，為何會選《紐約時報》而非其他報紙？這是因為，這份報紙有著大量的專欄，可以代表各個時代報紙產業的趨勢。

自創立以來，紐約時報每天使用的圖表量

Though statistical graphics were already in fairly common use by the turn of the 19th Century, the pre-1930 editions we examined had exactly zero charts. Every paper since 1930 has had at least one chart, and the 2010 and 2015 editions each included more than twenty.

雖說早在19世紀末期，已經有人開始使用圖表，但我們發現一直到1930年以前，紐約時報幾乎沒有出現過圖表。到了1930年開始，每天才至少出現一張圖表，而到了2010年跟 2015年，每天都刊登超過20張圖表。

The chart is now accepted as an integral tool for journalism. What took so long?

現今大眾都接受報紙上出現圖表這種整合性的工具，但中間的過程到底花了多長的時間呢？

The Dark Ages (1930 and Before)   1930年以前：黑暗時期

 

The modern statistical chart was invented by the Scottish political economist William Playfair at the end of the 18th Century. In a fit of astonishing innovation from 1786 to 1801, Playfair made the first known bar chart, line chart, and pie chart.

講到圖表的起源，不得不提現代的統計圖表是由蘇格蘭經濟學家威廉普萊菲〈William Playfair〉在18世紀末期所發表的。光是在1786年到1801年間，普萊菲就製作了現今流傳於世的三張圖表：長條圖〈bar chart〉、折線圖〈line chart〉和圓餅圖〈pie chart〉。

Historians believe that Playfair’s development of the statistical chart was a response to the surge in the availability of economic and demographic data collected by governments. Suddenly, efficient ways to express all this information were necessary. The chart filled that need.

歷史學者相信普萊菲研製的統計圖表，是為了因應政府機關有著越來越多關於經濟與人口統計資料。在當時，極需要一種簡便的方式來傳達這些資訊，而普萊菲的圖表正好符合他們的需要。

Playfair’s data visualizations did not quickly catch on. They were initially considered childish by serious academics.

但很意外的是，現今風行這樣的資料視覺化，在當時並不像現在一樣有相當大的迴響。起初，被許多傳統派的統計學者認為，這樣的呈現方式是很幼稚的。

普萊菲於1786年研製的第一張折線圖

The use of statistical graphics did not proliferate in Playfair’s lifetime—he died in 1823. But by the beginning of the 20th Century, graphical representations of data were common in textbooks and used by governments, financiers, and scientists. Both the French and American governments annually published statistical atlases filled with charts that aimed to explain the world.

很遺憾，在普萊菲一生當中，一直到1823年它去世前，很少人知道這種偉大的發明。但到了20世紀初，在教科書上就經常看到這樣的圖表，而且經常被政府機關、財經專家和科學家拿來使用。法國和美國政府每年還會出版統計圖集，為的就是用簡便的方式傳遞給世人資訊。

Yet charts were nowhere to be seen in the newspaper.

儘管在當時，圖表就出現在非常多的地方，但在報紙上，並未發現它的蹤跡。

There were lots of statistics in the paper. Journalists wrote about trends in the bond markets, activity in the commodity markets, and changing birth rates. Numbers were a big part of the media; they just weren’t visualized.

在當時，的確有非常多的統計數據出現在報刊上。報社記者也已經開始撰寫了證券市場的趨勢、資本市場活動和生育率的變化這些現代常見的數值。統計數值早在當年就是報刊很重要的一部份，他們只是沒有統整成圖表而已。

This table from on cotton trade data from 1865 and story on birth rates from 1915 would certainly be candidates for visualization today.

左側的表格是1865年的棉花交易價格，而右側的內容是1915年的出生率和出生人口，這些在現代都能用圖表做成呈現，但在當年都是只用文字表現。

Given the vast amounts of data in the paper, why the rarity of charts?

既然有那麼多的資料統計出現在報刊上，那麼為何圖表會那麼少？

There is little academic research on the subject, but the most likely explanation is a combination of technological hurdles and the public’s unfamiliarity with statistical graphics.

雖然針對這個問題，並沒有科學上的證據，但最可能的原因不出兩種，首先是技術上的障礙，再者，當年大眾對於圖表其實是相當陌生的〈編按：當時的教育程度跟現在差很多〉。

Making charts in the first half of the 20th Century meant hand drawing them—with a little help from primitive technology. Chart makers' main tools were rulers, graph paper, and compass pens. Journalists did not make charts without seriously considering whether it was worth the time.

什麼是技術障礙呢？在20世紀的前半期，當時沒有美工軟體還有先進的印刷科技。如果要放圖表在報刊上，那麼就需要親手繪製。以當時的工具來說，只有尺、製圖紙和圓規而已，所以無法製作出像現代那樣精美的圖表，而且製作圖表的程序也相當繁複，無庸置疑地，報社記者並不會考慮要在報刊上加入圖表-它得花上非常多的時間〈編按：編輯文字時間都不夠了，還來整理圖表〉

When newspapers did choose to publish a chart, the average person might not understand it. Though data visualization was common in academia, government, and science, the average person was not accustomed to interpreting charts. Charts were still such a rarity in the early part of the 20th Century that in 1913 there was a “parade of statistical graphics” in New York City .

另外，在當時如果報紙真的放上了圖表，大部分的人其實可能沒辦法了解。雖然當時資料視覺化在學界、政府和科學就已經相當盛行，但是在20世紀早期，民眾看到圖表的機會依舊是相當少的，甚至在1913年，學界為了要推廣圖表，在紐約市還辦了一場〝統計圖表大遊行〞，利用遊行的過程將圖表展示給大眾。

Charts were such a novelty in 1913 that there was a parade of statistical graphics.

在20世紀初期，圖表對於大眾來說是一項相當新奇的事物，甚至在1913年有了〝統計圖表大遊行〞

The Business Chart Era (1930s-1970s)  1930-1970：商業圖表時代

July 4, 1933, is arguably the day that data visualization became a permanent part of mainstream media.

1933年7月4號，對於許多人來說，並不是一個特別的一天〈如果說特別，可能就是美國國慶日吧〉，但正是圖表開始變成主流媒體重要的一環。

On that day, with no fanfare, our analysis suggests that the New York Times published a stock market line chart for the very first time. A stock market line chart is that most famous of graphics that shows trends in the performance of an index of stocks over a given period.

從我們蒐集的資料可以推測，那天，正是歷史上，紐約時報放股市折線圖的第一天。從那天起，這樣的資料呈現就注定走上了最流傳於世的圖表的道路上，因為她是最簡潔地顯示每個階段每支股票的表現方式。

Charts did appear in the Times prior to that day, but inconsistently. Some days the paper contained no charts at all. But since July 4, 1933, the Times has published a data visualization almost every Tuesday through Saturday. (The markets are closed on Saturday and Sunday, so there is no chart the following day.) We also checked the archives of the LA Times and San Francisco Chronicle, and neither published a stock market line chart before July 4, 1933. The LA Times was also already periodically publishing charts.

其實在此之前，圖表就曾在《紐約時報》出現過，但並不是每天都有。有時候甚至連一張圖表都沒有出現過。但自1933年7月4號這天開始，《紐約時報》幾乎在每周二到六都放上了圖表。〈這是因為，股票市場在周末休市，所以在周日跟周一並不會放上圖表。〉同樣的在這一天前，我們也發現在《洛杉磯時報》和《舊金山紀事報》的專欄上，都沒有任何一張圖表在上面。

The first stock market line chart published in the New York Times was in 1933.

在1933年的7月4號這天，紐約時報首度刊登股市折線圖，同時也是第一份放上股市折線圖的報刊

Even after the Times began regularly publishing data visualizations, they were not spread evenly throughout the paper. From the 1930s to the 1970s, the preponderance of charts appeared in the Business section.

儘管在那天之後，紐約時報定期發布這樣的圖表，但從1930年到1970年間，圖表也只是出現在財經版，並沒有在其它版出現過。

More specifically, the vast majority of mid-century newspaper graphics were line charts showing commercial trends like changes in the stock market, the production of automobiles, or the amount of electricity produced. It’s probably no coincidence that these charts became common during the Great Depression—when the economic health of the United States weighed heavily on the minds of Times readers.

講得更仔細一點，在這個時期，大多數的報刊上都是刊登有關商業趨勢的折線圖，像是股市、汽車生產量還有電力生產量。這樣的現象可能是因為當時的經濟狀況，尤其是在經濟大恐慌〈1929-1933年間〉爆發後，有越來越多民眾重視相關的經濟資訊，於是這樣的圖表就越來越常出現在報紙上了。

This chart from a 1940 edition of the New York Times shows trends in the production of freight carloading. Most of the charts found in the New York Times in the 1930s to 1970s displayed economic and business trends.

這張圖表是源自於1940年的紐約時報，它所闡述的是在1929年至1940年間貨運裝載量的變化。在1930年至1970年間，出現在紐約時報的圖表多半是在闡述財經相關的趨勢。

The fact that the business section was the earliest home of the chart is not a surprise. Business and economics has always been at the forefront of statistical graphics, and the economy produces lots of statistical data. The business-focused Wall Street Journal was already publishing charts in the early 1920s.

其實，財經版是報刊最早出現圖表的版面並不會太意外。因為商業和經濟這兩個領域都有非常多的資料，於是就有了資料整合的需求，而且經濟學界也製作非常多的圖表。就連華爾街日報，這份以商業和經濟這兩個領域聞名於世的報刊，早在1920年代初期就已經放上了統計圖表。

A chart from a 1921 edition of the Wall Street Journal.

1921年刊載於華爾街日報的一張圖表

Even today, most Times’ charts are in the Business section—18 of 25 for the September 16, 2015 edition. This phenomenon can be observed across the media spectrum (i.e. The Economist publishes a lot more charts than The New Yorker).

時至今日，大部分在紐約時報上的圖表都是出現在財經版，以2015年9月16號那天的紐約時報為例，那天的報紙一共出現25張圖表，其中竟然有多達18張圖表出現在財經版上。而這樣的現象不只出現在紐約時報，在整個媒體業都是一樣的情況。〈例如：《經濟學人》這份雜誌裡的圖表就比紐約時報還要多〉

While most of the graphics in the Times during this period were simple line charts, there were some more complex visualizations. The following chart from a 1960 edition of the Times is an early version of a lattice graphic—side by side displays of the same information for different groups. Inventive charts like this one were a sign of things to come.

儘管這段時期多數的圖表仍只是簡單的折線圖，但也開始越來越多相對複雜的圖表在這個時期出現。就像是下面的圖表，正是出現在1960年的紐約時報，它算是早期的組合圖〈就是現在微軟的文書軟體裡出現的圖表，結合折線圖跟柱狀圖的圖表〉，它所呈現的是同個條件下，不同類型的呈現。像這樣在當時前所未有的圖表出現，證明了圖表正往下個階段邁進。

Most early charts in the newspaper were simple line charts, but more complex graphics, like this one from 1960, sometimes appeared.

圖表出現在報紙的初期，多半只是簡單的折線圖，但是偶而會出現像是1960年這張較為複雜的圖表。

Computerization and the Rise of the Non-Business Chart (1980s-2000s)

1980-2000：電算化時代及非經濟圖表的興起

The seeds of the newspaper chart explosion at the end of the 20th century were sewn in the offices of IBM and Bell Labs decades earlier.

在IBM和貝爾實驗室興盛的開始，就是埋下了在20世紀晚期，報紙圖表數爆量增長的種子。

In the 1950s and 60s, engineers developed computer programming languages for the processing of statistical data (FORTRAN and S) and programs that simplified data visualization (SAS and Lisp-Stat). These new tools for data analysis and visualization were only comprehensible to a small number of programmers. By the 1980s, however, software programs—like Lotus 1-2-3 and Microsoft Excel—were created and allowed more people to easily make charts. Instead of hand drawing charts or using complicated graphic terminals, journalists could now produce charts on personal computers.

在1950和60年代，IBM與貝爾實驗室工程師研發了處理統計數據的電腦程式語言〈FORTRAN 福傳跟S語言〉還有可以簡化資料視覺化的程式〈SAS 統計分析系統和Lisp-Stat〉。在當時，這些用於資料分析和視覺化的新工具，由於操作不易，也只有像是程式設計師這樣小眾的人口在作使用。但是到了1980年代，像是Lotus 1-2-3〈1983年推出的電子試算表軟體〉和Excel〈1985年推出電子試算表軟體，於1993年擊敗前者〉這樣的程式軟體推出後，讓人們更簡便地製作圖表了。也因此，報社記者不用再像以前一樣，辛苦地用手繪製圖表或是使用複雜的圖像終端程式〈graphic terminals〉，自這個時期開始，就只要使用個人電腦就可以製作了。

Before personal computers, newspapers used graphics terminals to make charts.

在個人電腦尚未問世之前，報社只能用圖像終端程式來製作圖表

From 1950 to 1975, the newspapers we examined averaged only two charts per edition. From 1980 to 2005, this number increased to over 10.

從1950到1975年間，這個階段我們所收集到的報紙每一天平均都只出現2份圖表。而1980至2005年間，則是增加到10份以上。

The look of the charts in the Times also changed with computerization. The following chart from a 1980 issue of the Times has a similar crispness to modern data visualization.

拜電算化〈computerization〉所賜，在紐約時報上的圖表樣式也有了變化。就像是下方的那張圖表，正是出現在1980年的紐約時報上，是不是很像現在常見的圖表呢？

This chart from 1980 looks similar to what we find in papers today.

這份圖表是出現在1980年的紐約時報，它就像是現在我們常見的圖表。

Data visualization sophisticates might notice that the bar chart above exhibits one of the great faux pas of good data visualization: The y-axis begins at 380, rather than at zero. This misleads the undiscriminating reader about the magnitude of the trend displayed by the chart.

對於圖表相當了解的人可能會發現下方的圖表正是一個最失敗的案例：它的y軸起始點是在380，而非我們常見的0。這樣的呈現會誤導沒有仔細看圖表的讀者。

In the 1980s and 1990s, the Times was full of examples of what is now considered poor chart making. For example, the 1995 chart below shows that the Times was not immune to the peculiar desire of many chart makers to make their visuals 3-D (which obscures the reader’s ability to comprehend the data).

除了這張圖表外，紐約時報在1980至1990間也做出非常多在現在看來非常不好的圖表，就以下方的1995年的圖表為例，紐約時報的確繪製出相當有創意的3D圖表，但這讓讀者很難理解這份圖表所代表的意義。

The New York Times was not immune to making their charts 3-D and thus more challenging to read.

這份圖表要表達的是1995年六月到九月，每天美金兌換日圓的匯率變化，的確這是一份極具創意的圖表，但是這份圖表對於許多民眾來說真的很難看懂，除了這樣立體圖表的呈現會誤導數值外，尤其當您把視線放在圖表最右側時，您會很看得出來每一格的數字〈編按：其實如果要呈現上面的資訊，只要用一般的折線圖就可以了〉

As charts became easier to make and more widely understood, their use spread throughout the paper.

當圖表越來越容易製作和理解時，它就出現在報紙上的頻率順帶增加，甚至會讓讀者覺得，報紙如果沒有圖表會很奇怪。

In our sample of Times pieces, we only found one chart used outside the Business section before 1990. But starting in 1990, each of the issues we checked had at least one chart outside the Business section. For example, the September 13, 1990, edition of the Times had the following charts in the Metro and Health sections.

講完了圖表的呈現，那麼我們再回到報紙版面。前段有提到，在1930至1970年間，圖表只出現在財經版，雖然圖表製作越來越簡便和多元，但是一直到1990年以前，我們發現只有一張圖表出現在其他版上。不過，到了1990年後，我們發現每天至少會有一張圖表出現在非經濟版上。以1990年9月13號當天的報紙為例，在市政與康健版〈metro and health，這個版好像現在已經不存在〉上有出現下方這樣的圖表。

Starting in the 1990s, charts were regularly found outside the business section.

從1990年開始，圖表不再專屬於財經版。

The Modern Era (Late 2000s to Today)  2000末期到現在：現代

By the late 2000s, charts had become central to the journalism at the Times. The newspaper had a team of graphics reporters infusing the paper with innovative visualizations. For the Times web edition, those data visualizations were often interactive.

到了千禧年代的末期，圖表的地位已經搖身一變，成為紐約時報的重心，甚至有更多元的圖表設計，報社還組織一個圖表團隊，裡面成員都是專精於繪製圖表的記者，現在看到很多創新的圖表，都是由這個團隊所研發的。

The following “stained glass window” style chart from 2008 shows how sophisticated visual journalism became. The chart displays inflation over the past year by expense types. Web readers could hover over different parts of the image to get additional information.

就像是下方〝花窗玻璃〞風格的圖表，正是在2008年研製出來的，就可以想見現今的視覺化報導竟然到了這樣的複雜程度。這張圖表是在講述過去消費者在各類支出的成長率。習慣上網的讀者如果不滿意顏色的呈現，甚至可以把裡面的顏色作修改。

By 2008, the data visualization in the New York Times was complex and interactive.

早在2008年，資料視覺化就已經變得如此複雜和多元。

The Times has been at the vanguard of making complex statistical graphics, but the Washington Post, USA Today, and Los Angeles Times have also committed significant resources to data visualization over the last decade.

在複雜統計圖表的設計上，紐約時報已是這方面的翹楚，但在過去十年，包括《華盛頓郵報》、《今日美國》和《洛杉磯時報》也開始設計出多樣化的圖表。

Not only did newspaper charts become more complex, they became more common. The September, 14, 2010, edition of the New York Times had 26 different charts, seven of which were outside the business section. The following image shows every chart in that print edition of the paper.

報紙圖表的形態越來越多元外，也越常出現在報紙上。在2010年9月14號這天的紐約時報，就出現了26種不同的圖表，其中有7個是出現在財經版以外的版面。下面的圖片是那天出現在報紙上的所有圖表。

The New York Times cemented their prioritization of visual journalism when they launched the data-driven news site the Upshot in 2014. The Upshot’s original team of fifteen included three full-time graphic journalists. Amanda Cox, the site’s current editor, started at the Times in 2004 as a graphics reporter.

到了2014年，紐約時報為了要讓資料視覺化的技術更精進，推出了一個名為Upshot的網站，是以資料統整為主的新聞網。Upshot的草創團隊一共有15位，其中有三位是全職的圖像記者〈graphic journalist〉。而亞曼達考克斯〈Amanda Cox〉這位現在是網站的編輯，在投身網站編輯前，正是其中一位圖像記者。

Today, it is not unusual for an Upshot graphic to occupy the most prominent space on the front page of the New York Times website. Readers checking the site on the morning of September 9, 2016, were greeted by a variation on a bar chart comparing the growth in murder rates across cities.

現在我們在紐約時報官網封面上看到的那些精美的圖像正是由Upshot所繪製的。在2016年9月9號的開始，當紐約時報開始放上由Upshot所製作的第一張條狀圖時，讀者都相當滿意這樣的呈現。

The front page of the New York Times website on the morning of September 9, 2016 was dominated by a statistical graphic.

2016年9月9號早晨的紐約時報官網首頁。大多數的讀者都注意到了這張圖表。

Over the course of the last century, the chart has gone from obscure academic tool to a mainstay of the front page. This revolution in data visualization came about thanks largely to better technology and more accessible data. These days, a newspaper without a chart would barely feel like a newspaper at all.

過去這一百年來，圖表已經從難解的學術工具變成網站首頁的支柱。這樣的轉變得歸功於科技的進步和更多可用的資訊。圖表的普及讓我們很難想像沒有圖表的報紙會是什麼樣子。

原文：When Did Charts Become Popular

原文作者：Dan Kopf

撰文日期：2016.9.14