TOP TRENDING

Speeches by US politicians "are about the reading age of a 13-year-old"


Congressional speeches by US politicians have become easier since the 1970s and, according to one study, only require the reading age of a 13-year-old.

Kansas State University computer scientists analyzed two million congressional speeches by Republican and Democratic leaders delivered from early 1873 through late 2010 – a total of 183 years of speeches.

Text analysis algorithms were used by the researchers to find out how congressional speeches have changed in terms of complexity, emotion, and divisibility.

Newer speeches use smaller vocabulary, simpler language, and talk more about "the other party" than speeches given a decade ago, the authors found.

The researchers compared the vocabulary in the speeches with the average reading levels in US schools – for example, a 13-year-old would be in eighth grade or ninth grade in the UK.

The authors attributed the decline in reading level to the surge in the broadcast media in Congress that began in the mid-1970s – with politicians "playing in front of the camera."

The study comes out when Joe Biden is officially named the Democratic nominee for president in dozens of speeches, including former President Bill Clinton who called on President Trump to "watch TV for hours".

Kansas State University computer scientists analyzed two million Congressional speeches by Republican and Democratic politicians delivered between 1873 and 2010

This table assigns a school grade reading ability (x-axis) to speeches made by US politicians and shows a decrease in the level required to understand the speeches over the past 40 years

This table assigns a school grade reading ability (x-axis) to speeches made by US politicians and shows a decrease in the level required to understand the speeches over the past 40 years

The number of speeches in Congress varies, but researchers found they are on an upward trend – with almost twice as many in the 2000s as in the 1990s.

The reading comprehension of speeches has changed significantly over the years – complexity first increased and became less complex since the 1970s.

During the analysis, the Coleman-Liau readability index was measured, which estimates the reading status of a certain text and links it to the corresponding school grade.

The analysis found that the reading level of congressional speeches by both Republican and Democratic lawmakers increased steadily from eighth grade in the 19th century to tenth grade in the 1970s.

But since 1976 the reading level of political speech has steadily declined and from the 21st century is below the reading level of the ninth grade.

The same trend has also been seen in the vocabulary used by members of Congress in speeches, which steadily increased and then decreased until the early 1970s – and it's still declining, said co-author Lior Shamir.

President Barack Obama

President George W. Bush

During the George W. Bush administration, speeches by Democratic lawmakers expressed more negative sentiments compared to their Republican counterparts, and they turned around when President Barack Obama took office

According to the study, the decline in the reading level and vocabulary of the speeches may be related to the increasing media presence.

This includes live radio and television coverage of the Congress from the 1970s.

Members of Congress began gradually adapting their speaking styles and reaching the public through the media instead of reaching out to their peers.

As part of the Congressional Speech Study, the team, including students Ethan Tucker and Colton Capps and Professor Shamir, also looked at the mood.

The algorithms measured aspects of the speeches such as vocabulary, reading level and the positive or negative feelings expressed in the speeches.

"Based on this analysis, the algorithm determines whether a piece of text is positive, very positive, negative, very negative, or neutral," said Shamir.

The algorithms also measured the frequency with which various topics were discussed, to find out how often ideas were repeated based on the thousands of speeches given in Congress each year.

Research showed that the frequency of words relating to the identity of women – like her, her, hers, wife, women – has steadily increased since the early 1980s, while the frequency of words that identify men has decreased Has.

The frequency of words relating to women's identity is five times higher in the 21st century than it was in the 1950s, but lower than that of words relating to men's identity.

Since the 1990s, women's identity terms have been more common in speeches by Democrats than in speeches by Republican lawmakers.

The researchers found that the number of speeches rose sharply in the 2000s, almost twice what it had been in the past decade

The researchers found that the number of speeches rose sharply in the 2000s, almost twice as much as in the last decade

"For most of the 20th century, however, there were no significant differences between women's identities in Democratic and Republican speeches, and expressions of women's identity were about ten times less common than expressions of male identity by lawmakers from both parties," Shamir said.

Analysis of the speeches by the researchers also showed that recent congressional speeches expressed more positive and negative feelings than speeches given in Congress in the 19th and early 20th centuries.

The mood in political speeches gradually became more positive and peaked in the 1960s, but declined sharply in the 1970s.

Since the 1970s, the feelings expressed in congressional speeches have become more positive again, the authors noted.

Using text analysis algorithms, it examined how congressional speeches have changed in terms of complexity, emotion and division in 138 years

Using text analysis algorithms, it examined how congressional speeches have changed in terms of complexity, emotion and division in 138 years

Another aspect that was reflected in the analysis was the split among the partisans, Shamir said.

From the mid-1990s, Republican and Democratic speeches became increasingly divergent and correlated with the White House party.

During the George W. Bush administration, speeches by Democratic lawmakers expressed more negative sentiments than their Republican counterparts.

That difference reversed immediately after 2008, with the beginning of the Obama administration, during which Republican speeches became more negative.

The study only looked at speeches up until 2010, so there is no reference to Donald Trump or the legislature's response to the extremely divisive president.

The results were published in Heliyon magazine.

HOW TO LEARN ARTIFICIAL INTELLIGENCES WITH NEURAL NETWORKS

AI systems are based on Artificial Neural Networks (ANNs) that try to simulate how the brain works in order to learn.

ANNs can be trained to recognize patterns in information – including speech, textual data, or visual images – and are the foundation of a large number of developments in AI in recent years.

Traditional AI uses input to “teach” an algorithm about a particular topic by feeding it vast amounts of information.

AI systems are based on Artificial Neural Networks (ANNs) that try to simulate how the brain works in order to learn. ANNs can be trained to recognize patterns in information - including speech, textual data, or visual images

AI systems are based on Artificial Neural Networks (ANNs) that try to simulate how the brain works in order to learn. ANNs can be trained to recognize patterns in information – including speech, textual data, or visual images

Practical uses include Google's voice translation services, Facebook's face recognition software, and Snapchat's live image manipulation filters.

Entering this data can be very time consuming and is limited to one type of knowledge.

A new generation of ANNs called Adversarial Neural Networks pit the minds of two AI bots against each other, allowing them to learn from each other.

This approach aims to accelerate the learning process and refine the results generated by AI systems.

(tagsToTranslate) dailymail (t) sciencetech