Speech sounds are the basic units of human language, and analyzing them can reveal a lot about how we communicate, learn, and vary across cultures and contexts. But how can you study speech sounds effectively and accurately? In this article, you will learn about some of the best tools for analyzing speech sounds, from software to hardware, and from acoustic to articulatory.
One of the most widely used software tools for analyzing speech sounds is Praat, a free and open-source program that allows you to record, edit, visualize, and measure speech sounds. Praat can display waveforms, spectrograms, pitch contours, intensity curves, and formant tracks, as well as perform various types of analysis, such as segmentation, annotation, synthesis, manipulation, and statistical testing. Praat is also compatible with many scripting languages, such as Python and R, and can be integrated with other software, such as ELAN and Audacity.
Software like Praat allows from undergraduate students to post-doctoral students to dive into speech analysis. During my undergraduate studies, I enjoyed using the tool. I wish I could use it more in the future and learn to integrate it with Python and R, as the description suggests here.
Listen to as many listening passages as you possibly can. Follow as many channels as you can. Concentrate on the things you enjoy
Another important aspect of analyzing speech sounds is the hardware that you use to capture and produce them. Depending on your research goals and budget, you may need different types of hardware tools, such as microphones, headphones, speakers, sound cards, and audio interfaces. Some of the factors that you should consider when choosing hardware tools are the quality, sensitivity, frequency range, noise level, and compatibility of the devices. You should also make sure that your hardware tools are calibrated and adjusted properly to ensure reliable and valid results.
Select your hardware tools, taking your research participants into account. If you are analysing speech sounds of a small child, make sure that your recording equipment does not scare them or make them uncomfortable. If you are visiting your research participants outside the lab (either due to their health condition, or because it is important for you to record them in a setting that is more natural to them), make sure that your equipment is easily portable,
Acoustic tools are those that focus on the physical properties of speech sounds, such as frequency, amplitude, duration, and spectrum. Acoustic tools can help you to describe and compare speech sounds in terms of their acoustic features, such as pitch, loudness, length, and quality. Some examples of acoustic tools are the harmonic-to-noise ratio (HNR), which measures the degree of periodicity or noise in a speech sound; the voice onset time (VOT), which measures the time between the release of a stop consonant and the onset of voicing; and the spectral tilt, which measures the slope of the energy distribution across frequencies.
Articulatory tools are those that focus on the physiological aspects of speech production, such as the movements and positions of the vocal tract organs, such as the tongue, lips, jaw, and larynx. Articulatory tools can help you to understand and model how speech sounds are produced and modified by the speakers. Some examples of articulatory tools are the electroglottograph (EGG), which measures the contact between the vocal folds; the ultrasound, which images the shape and movement of the tongue; and the electromagnetic articulograph (EMA), which tracks the motion of sensors attached to the articulators.
Perceptual tools are those that focus on the psychological aspects of speech perception, such as how listeners identify, categorize, and interpret speech sounds. Perceptual tools can help you to investigate how speech sounds are processed and affected by various factors, such as context, background noise, speaker characteristics, and listener expectations. Some examples of perceptual tools are the identification task, which asks listeners to label speech sounds based on a given set of options; the discrimination task, which asks listeners to judge whether two speech sounds are the same or different; and the rating task, which asks listeners to evaluate speech sounds based on a given scale or criterion.
Finally, ethical tools are those that help you to conduct your speech sound analysis in a responsible and respectful way, taking into account the rights, interests, and well-being of the speakers, listeners, and communities involved in your research. Ethical tools can help you to address issues such as informed consent, confidentiality, data management, dissemination, and impact. Some examples of ethical tools are the code of ethics, which outlines the principles and standards that guide your research conduct; the consent form, which informs and obtains the agreement of the participants; and the data protection policy, which specifies how you store, share, and delete your data.